The role of Broca`s area in regular past-tense morphology

Transcription

Neuropsychologia 49 (2011) 1–18
Contents lists available at ScienceDirect
Neuropsychologia
journal homepage: www.elsevier.com/locate/neuropsychologia
The role of Broca’s area in regular past-tense morphology:
An event-related potential study
Timothy Justus a,∗ , Jary Larsen a , Jennifer Yang a , Paul de Mornay Davies b ,
Nina Dronkers a,c , Diane Swick a,c
a
b
c
Medical Research Service, VA Northern California Health Care System, Martinez, CA, USA
Middlesex University, London, UK
University of California, Davis, USA
a r t i c l e
i n f o
Article history:
Received 29 January 2010
Received in revised form 29 August 2010
Accepted 20 October 2010
Available online 28 October 2010
Keywords:
Aphasia
Broca’s area
Left inferior frontal gyrus (LIFG)
Event-related potentials (ERP)
Inflectional morphology
Regular and irregular past tense
a b s t r a c t
It has been suggested that damage to anterior regions of the left hemisphere results in a dissociation
in the perception and lexical activation of past-tense forms. Specifically, in a lexical-decision task in
which past-tense primes immediately precede present-tense targets, such patients demonstrate significant priming for irregular verbs (spoke–speak), but, unlike control participants, fail to do so for regular
verbs (looked–look). Here, this behavioral dissociation was first confirmed in a group of eleven patients
with damage to the pars opercularis (BA 44) and pars triangularis (BA 45) of the left inferior frontal gyrus
(i.e., Broca’s area). Two conditions containing word-onset orthographic–phonological overlap (bead–bee,
barge–bar) demonstrated that the disrupted regular-verb priming was accompanied by, and covaried
with, disrupted ortho-phonological priming, regardless of whether prime stimuli contained the regular inflectional rhyme pattern. Further, the dissociation between impaired regular-verb and preserved
irregular-verb priming was shown to be continuous rather than categorical; priming for weak-irregular
verbs (spent–spend) was intermediate in size between that of regular verbs and strong verbs. Such continuous dissociations grounded in ortho-phonological relationships between present- and past-tense forms
are predicted by single-system, connectionist approaches to inflectional morphology and not predicted
by current dual-system, rule-based models. Event-related potential data demonstrated that N400 priming effects were intact for both regular and irregular verbs, suggesting that the absence of significant
regular-verb priming in the response time data did not result from a disruption of lexical access, and
may have stemmed instead from post-lexical events such as covert articulation, segmentation strategies,
and/or cognitive control.
© 2010 Elsevier Ltd. All rights reserved.
1. Introduction
A commonly held view in neuropsychology argues that the
non-fluent varieties of aphasia associated with damage to Broca’s
area result in difficulties with the production and perception of
regular past-tense forms, with irregular past-tense forms largely
preserved. From the perspective of generative linguistics, this supposed dissociation represents the disruption of a morphological
algorithm that (for English speakers) adds or strips the suffix
|-d|, realized as one of three different allomorphs (/d/, /t/, or
/Id/), to or from the present-tense form of regular verbs (e.g.,
∗ Corresponding author at: Cognitive Neuropsychology and Electrophysiology
Laboratory, VA Northern California Health Care System, 150 Muir Road, Research
Building 4, Martinez, CA 94553-4668, USA.
E-mail address: [email protected] (T. Justus).
0028-3932/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.neuropsychologia.2010.10.027
looked–look). In contrast, the irregular verbs, which do not fit this
pattern (e.g., spoke–speak), are represented lexically according to
such dual-system perspectives (e.g., Marslen-Wilson & Tyler, 1998;
Pinker, 1999; Pinker & Ullman, 2002). A recent dual-system proposal argues that the phonological forms of the regular past tense
automatically trigger an attempt at morphological segmentation
in any word form containing this “inflectional rhyme pattern”
(IRP) (Marslen-Wilson, 2007; Marslen-Wilson & Tyler, 2007; Post,
Marslen-Wilson, Randall, & Tyler, 2008). Alternatively, from a connectionist perspective such a dissociation reflects a disruption
of phonological processes within a single system that maps the
joint phonological and semantic relationships between morphologically related words. Regular past tenses, which overlap heavily
with, and share a straightforward mapping to, the corresponding
present-tense forms, are more greatly disrupted due to phonological deficits according to such single-system perspectives (e.g., Bird,
Lambon Ralph, Seidenberg, McClelland, & Patterson, 2003; Joanisse
2
T. Justus et al. / Neuropsychologia 49 (2011) 1–18
& Seidenberg, 1999; McClelland & Patterson, 2002). The nature and
robustness of such regular–irregular dissociations following damage to Broca’s area is the focus of the current research.
1.1. Dissociations in the production of regular and irregular
past-tense forms
One way in which regular and irregular past tenses have been
argued to dissociate is in overt production tasks. For instance, the
participant would be asked to repeat or read a past-tense form, or
to generate the past tense from spoken contexts such as “Every
day I walk; yesterday I
.” Ullman et al. (1997, 2005) argued
that aphasics with anterior/frontal damage, along with people
with Parkinson’s disease and Huntington’s disease, have greater
difficulty with regular past-tense forms on these tasks, because
of procedural, rule-based deficits. In contrast, they argued that
aphasics with posterior/temporal damage, along with people with
Alzheimer’s disease, have greater difficulty with irregular pasttense forms on these tasks, because of declarative, lexical deficits.
The nature of this dissociation was challenged, however, by
Bird et al. (2003, Exps. 1–2), who argued that the deficit in
frontal aphasics is not based in morphological rules but rather
phonological complexity. They tested a group of ten non-fluent
aphasics who showed greater problems with regular verbs on generation tasks and found that the dissociation diminished when
the consonant–vowel (CV) structure of the regular and irregular past-tense forms was matched (also see Braber, Patterson,
Ellis, & Lambon Ralph, 2005; Lambon Ralph, Braber, McClelland,
& Patterson, 2005).
Not only has the interpretation of the regular–irregular production dissociation been questioned, it is also unclear whether the
dissociation is reliable following frontal lobe damage. Only ten of
the fifty non-fluent aphasics screened by Bird et al. (2003) fitted the
pattern predicted by Ullman and colleagues. Further, in a metaanalysis of seventy-five non-fluent aphasics, Faroqi-Shah (2007)
found no systematic dissociation between regular and irregular
verbs on these production tasks. This is consistent with the performance of our patients on production tasks using both the screening
and CV-balanced stimuli of Bird and colleagues (Larsen, Justus, de
Mornay Davies, & Swick, 2008).
Patients with more posterior, fluent forms of aphasia sometimes
show irregular past-tense production deficits and preserved performance on regulars (Ullman et al., 1997, 2005). Together, such
observations arguably constitute a double dissociation between
regular and irregular past-tense morphology in anterior and
posterior aphasics. Whether these deficits in the production of
irregular past-tense forms arise from damage to the mental lexicon or instead to semantic processes constitutes a second debate
between dual-system and single-system perspectives (Miozzo &
Gordon, 2005; Miozzo, 2003; Patterson, Lambon Ralph, Hodges, &
McClelland, 2001; Tyler et al., 2004).
The majority of the neuroimaging studies that examined the
similarities and differences between regular and irregular morphology also used either overt or covert production of the past tense
– and in some cases noun plurals – in English, German, or Spanish (Beretta, Campbell, et al., 2003; de Diego Balaguer et al., 2006;
Desai, Conant, Waldron, & Binder, 2006; Dhond, Marinkovic, Dale,
Witzel, & Halgren, 2003; Jaeger et al., 1996; Joanisse & Seidenberg,
2005; Sach, Seitz, & Indefrey, 2004; Sahin, Pinker, & Halgren, 2006;
for discussions see Beretta, Carr, Huang, & Cao, 2003; Jaeger, 2003;
Jaeger, Van Valin, & Lockwood, 1998; Seidenberg & Arnoldussen,
2003; Seidenberg & Hoeffner, 1998). Like the patient meta-analysis
of Faroqi-Shah (2007), these studies also failed to support an
anterior–posterior distinction between regular and irregular verb
morphology. Generally speaking, the generation of regular and
irregular past-tense forms tends to recruit similar cortical networks
when compared to reading the corresponding present tense. Further, the majority of these studies demonstrated greater inferior
frontal involvement for the production of the irregular past, in contrast to the predictions of the declarative-procedural model.
The interpretation of these patterns differed widely among
the authors of these studies, with some emphasizing the general
similarity of regulars and irregulars (Desai et al., 2006; Dhond
et al., 2003; Sach et al., 2004), and attributing the greater activity for irregulars to the single-system notion of less consistent
phonological mapping between present- and past-tense form
(Seidenberg & Hoeffner, 1998; Seidenberg & Arnoldussen, 2003).
Others interpreted any differences between regulars and irregulars as consistent with dual-system approaches (Beretta, Campbell,
et al., 2003; de Diego Balaguer et al., 2006; Jaeger et al., 1996), or
focused on the role of the left inferior frontal gyrus (LIFG), arguing
for a rule-like grammatical function for this region that would seem
to apply equally to both regulars and irregulars (Sahin et al., 2006).
1.2. Dissociations in the perception of regular and irregular
past-tense forms
The present work follows from a related literature arguing for a
dissociation, following damage to the frontal lobe, between regular
and irregular past tenses in tasks of perception and lexical access.
Marslen-Wilson and Tyler (1997) presented data from a priming
task in which past-tense primes immediately preceded presenttense targets, with a lexical decision for the target item. In this
study, the two aphasic patients showed priming for irregular items
(e.g., bent–bend) but inhibition for regular items (e.g., baked–bake),
whereas controls and a right hemisphere patient showed priming for both. Tyler, de Mornay Davies, et al. (2002, Exp. 1) tested
the same two aphasic patients and three additional aphasics in a
similar experiment. In this case, the patients as a group demonstrated priming for the irregular items and a flat effect for the
regular items, whereas controls primed significantly for both, and
primed significantly more for the regular items compared to the
patients. Both groups demonstrated significant semantic priming
(e.g., cherry–grape) but neither demonstrated significant priming
for word-onset ortho-phonological overlap (e.g., gravy–grave), also
known as formal (form-based) priming. In an extension of their
work with these patients, Longworth, Marslen-Wilson, Randall,
and Tyler (2005) found that priming from a past-tense form to
a semantically related present-tense form was disrupted for regular pasts (e.g., blamed–accuse) but not for irregular pasts (e.g.,
shook–tremble). Marslen-Wilson and Tyler (2007) argued that the
lack of priming between regular past- and present-tense forms
represents a failure to decompose the morphologically complex
regular past tense into stem and affix, resulting in a disruption of
lexical access for the corresponding present-tense form.
It is important to note that in studies employing the immediatepriming design, no patient group consistently demonstrates the
reverse pattern: preserved priming for regulars and a lack of priming for irregulars. Thus, the single dissociation represented by the
lack of regular past-tense priming may simply represent a failure
to observe a less robust effect (regular-verb priming) due to the
additional variability and statistical noise, whereas a more robust
effect (irregular-verb priming) continues to be observed. Indeed,
one notable aspect of this experimental design is that in healthy
controls, irregular verbs consistently prime numerically, and sometimes significantly, more strongly than regular verbs do (Justus,
Larsen, de Mornay Davies, & Swick, 2008; Justus, Yang, Larsen,
de Mornay Davies, & Swick, 2009; Longworth, Keenan, Barker,
Marslen-Wilson, & Tyler, 2005; Marslen-Wilson & Tyler, 1997;
Tyler, de Mornay Davies, et al., 2002; Tyler et al., 2004). This pattern
is not to be confused with a different dissociation that is sometimes
found when there are intervening items between prime and tar-
get. In such delayed-priming designs, it has been argued that only
regular pasts prime the corresponding present-tense form (Napps,
1989; Sonnenstuhl, Eisenbeiss, & Clahsen, 1999; Stanners, Neiser,
Hernon, & Hall, 1979; but see Fowler, Napps, & Feldman, 1985;
Hanson & Wilkenfeld, 1985).
In the current study, we revisited the dissociation between regular and irregular past-tense priming in anterior aphasics reported
by Marslen-Wilson and Tyler (1997) and Tyler, de Mornay Davies,
et al. (2002), asking the following four questions.
1.2.1. Does the priming of regular and irregular verbs dissociate
following damage to Broca’s area?
First, we asked whether this behavioral dissociation – a failure to
demonstrate significant priming between regular past-tense forms
and present-tense forms, with preserved priming for irregulars –
would replicate in a larger, more homogeneous patient group than
had been previously reported. We included eleven patients with
confirmed damage to Broca’s area, operationally defined as the
pars opercularis (BA 44) and pars triangularis (BA 45) of the left
inferior frontal gyrus (LIFG) (see Keller, Crow, Foundas, Amunts, &
Roberts, 2009). Tyler, de Mornay Davies, et al. (2002) focused on
the role of the LIFG when interpreting the regular–irregular priming dissociation, even though the five patients in that study had
extremely diverse left-hemisphere lesions extending far beyond
the LIFG into temporal and parietal cortex. A stronger link between
disrupted regular-verb priming and damage to the LIFG specifically
was shown in a study of 22 brain-lesioned patients who were unselected for lesion location; in this group, tissue density in the LIFG
predicted the size of regular-verb priming effects (Tyler, MarslenWilson, & Stamatakis, 2005). The present study complements this
work by testing regular- and irregular-verb priming in a group of
lesion patients who were specifically selected based on damage to
the LIFG.
Both dual-system and single-system theories can explain the
dissociation between regular- and irregular-verb priming following LIFG damage, but for different reasons. Dual-system approaches
(e.g., Pinker, 1999; Pinker & Ullman, 2002), including the IRP viewpoint of Marslen-Wilson and Tyler (2007), argue that the LIFG
mediates morphological segmentation, whereas the single-system
approach (e.g., Bird et al., 2003; Joanisse & Seidenberg, 1999;
McClelland & Patterson, 2002) argues that the LIFG contributes to
phonological processing more generally. Before attempting to distinguish these two interpretations of LIFG function, however, it was
essential to confirm that the regular–irregular priming dissociation
is in fact associated with damage to this brain region.
1.2.2. Is the regular–irregular dissociation related to phonological
factors, and if so, does the inflectional rhyme pattern have special
status?
Having documented a dissociation between regular- and
irregular-verb priming, we then considered the data in light
of competing predictions of the dual-system and single-system
approaches. One question that distinguishes the two approaches
concerns whether the regular–irregular priming dissociation is
related to word-onset ortho-phonological priming. Two additional priming conditions were included: a pseudopast condition
(bead–bee), which was designed to mimic the phonological relationship between regular past- and present-tense forms (i.e., the
inflectional rhyme pattern or IRP), and an orthophono condition
(barge–bar), in which prime and target were related by the removal
of another phoneme that did not signal a potential morphological relationship. Most dual-system approaches (e.g., Pinker, 1999;
Pinker & Ullman, 2002) argue that problems with regular morphology are due to a disruption of rule-based affixation, and do
not specifically predict any relationship between regular morphological priming and phonological priming. However, the IRP
3
viewpoint of Marslen-Wilson and Tyler (2007) predicts that regular pseudopast words such as bead will be parsed as potential
stems and affixes, given that their phonological forms follow the
inflectional rhyme pattern of the regular past tense. Thus, the IRP
approach predicts that reduced priming for regular verbs should
be accompanied by reduced priming for the pseudopast condition
alone. The single-system approach (e.g., Bird et al., 2003; Joanisse &
Seidenberg, 1999; McClelland & Patterson, 2002) argues that problems with regular morphology are based in phonology, but does
not explicitly differentiate between different kinds of phonological
relationships (such as those that cue morphological relationships
and those that do not). Thus, current single-system approaches
predict that reduced priming for regular verbs should be accompanied by reduced priming in both the pseudopast and orthophono
conditions.
The literature on phonological priming suggests that a combination of pre-lexical facilitation, lexical inhibition, and post-lexical,
strategic processes operate simultaneously in designs employing
word-onset overlap (McQueen & Sereno, 2005; see Justus et al.,
2009, for a review). However, Tyler, de Mornay Davies, et al. (2002)
did not observe significant phonological priming in controls using
the auditory immediate-priming design with lexical decision, making it difficult to evaluate their claims regarding the special status of
the inflectional rhyme pattern in this experimental design. Instead,
their argument regarding the IRP is based on a different experimental design involving same-different judgments (Post et al., 2008;
Tyler, Randall, & Marslen-Wilson, 2002; Tyler, Stamatakis, Post,
Randall, & Marslen-Wilson, 2005; but see Bird et al., 2003, Exp. 3).
However, our group has consistently observed phonological priming in controls with the former design (Justus et al., 2008), allowing
for these claims to be tested. One possibility for why Tyler and colleagues did not observe phonological priming is that this condition
was underpowered in their stimulus sets (i.e., there were only 24
phonological trials compared to the 48 trials used for morphological
conditions; our studies employ 50 items per cell in all conditions).
Other relevant factors include inter-stimulus interval (ISI), global
relatedness proportions, and the specific stimulus items chosen for
the two sets of studies.
1.2.3. Is the regular–irregular dissociation better described as
categorical or continuous?
Another question that distinguishes the dual-system and singlesystem approaches concerns whether the dissociation between
regular- and irregular-verb priming reflects a categorical distinction between regular and irregular morphology or a graded
continuum of morphological regularity. To address this, the irregular stimuli were separated into two additional categories: weak
irregular verbs, which are similar to regulars in that they end with
a dental stop consonant in the past tense (e.g., spent–spend), sometimes by affixation (e.g., slept–sleep), and strong verbs, which are
less similar to regulars, do not involve affixation and instead use
a system of vowel changes known as ablaut (e.g., spoke–speak).
This division allows us to ask whether any regular–irregular dissociation is categorical or continuous. Specifically, if the preserved
irregular-verb priming held equally for both kinds of irregularities,
this would suggest that the regular–irregular dissociation represents a categorical effect of regularity per se. This would be most
consistent with the dual-system approach, which argues that the
rule-based regular affixation pattern either is or is not applicable
(Pinker & Ullman, 2002). If, however, the preserved irregular-verb
priming were greater for the strong verbs compared to the weak
irregular verbs, this would suggest that the regular–irregular dissociation represents a continuous, graded effect (Justus et al., 2008,
2009). Marslen-Wilson and Tyler’s (2007) IRP viewpoint, while still
positing a categorical division between regular and irregular verbs,
might accommodate some degree of continuity in priming effects
4
given that about half of the weak irregular pasts in English contain
the inflectional rhyme pattern. Unlike both the traditional dualsystem and the IRP approaches, the single-system approach argues
that morphological regularity is inherently continuous, and thus
specifically predicts that dissociations between regular and irregular verbs will be graded, given that the phonological and semantic
differences between the two classes are a matter of degree (see
Joanisse & Seidenberg, 1999; Joanisse & Seidenberg, 2005; Kielar,
Joanisse, & Hare, 2008; Kielar & Joanisse, 2010).
1.2.4. Does the regular–irregular dissociation extend to
event-related potentials?
Finally, we asked whether the regular–irregular dissociation
could be captured using event-related potentials. In our previous
studies using this experimental design, priming between corresponding past- and present-tense forms resulted in reductions
of the N400, and in some cases, modulations of the late positive component, or LPC (Justus et al., 2008, 2009). The N400 is a
negative-going wave that peaks approximately 400 ms after stimulus onset. The component was first obtained in an experiment in
which sentences ended with semantically expected or unexpected
words (Kutas & Hillyard, 1980). However, the N400 is also sensitive to word repetition and semantic priming in word pairs (Rugg,
1985) and further can be used to distinguish individual words, in
the absence of context, along dimensions such as frequency (Van
Petten & Kutas, 1990), neighborhood density (Holcomb, Grainger,
& O’Rourke, 2002), and concreteness (West & Holcomb, 2000).
Accordingly, some authors consider the N400 to be an index of lexical access (Lau, Phillips, & Poeppel, 2008; Lau, Almeida, Hines, &
Poeppel, 2009) or of the access of conceptual knowledge associated
with a word (Kutas & Federmeier, 2000; Van Petten & Luka, 2006;
Federmeier, 2007), whereas others consider the component to be
an index of post-lexical processes, including semantic integration
(Hagoort, 2008).
The neural generators of the N400 component are believed to
include large portions of the left temporal lobe, with a smaller
contribution from the right, as evidenced by recordings of EEG
in patients, intracranial recordings, and magnetoencephalography
(Van Petten & Luka, 2006). Therefore, an overall disruption of the
N400 component was not expected in the LIFG-defined patient
group reported here (see Swick, Kutas, & Knight, 1998). However,
differential modulation of the N400 effect (i.e., the degree of difference between the N400s for unprimed and primed items) might be
predicted following LIFG damage, just as we might find a change
in the relative amount of time it takes to perform lexical decision
to primed versus unprimed items, despite no general deficit in the
task.
Such a difference would be predicted by the IRP theory of
Marslen-Wilson and Tyler (2007) as well as other dual-system theories that posit no whole-word lexical representation for regular
past-tense forms, but instead an LIFG-dependent affix-stripping
process that detects regular inflections, parses them, and then
inputs the stripped present-tense stem to the temporal lobe for lexical access. Within such a framework, LIFG damage would disrupt
lexical access for regular past-tense forms. This would be reflected
in the ERP data by a reduction in the N400 priming effect for regular
verbs. This reduction might be accompanied by a similar reduction
for the pseudopast condition, as predicted for the behavioral data.
The single-system perspective makes a similar prediction, but
for different reasons. The single-system theory argues that all pasttense forms are fully listed in the lexicon, and that past-tense
forms are mapped to the corresponding present tense based on
the interacting effects of ortho-phonological and semantic overlap. Accordingly, this approach might predict that N400 priming
effects in this design should reflect the interacting effects of orthophonological and semantic priming (see Kielar & Joanisse, 2010). If
damage to the LIFG disrupted priming based on ortho-phonological
overlap, and if these effects occurred prior to or during lexical
access, then this would also be reflected in the ERP data by a reduction in the N400 priming effect for regular verbs. This reduction
might be accompanied by similar reductions for both the pseudopast and orthophono conditions, as predicted for the behavioral
data.
The addition of ERP data to the experiment also permits us to
examine the events occurring after lexical access, as represented
by the N400, but before the participant provides a lexical decision response, often several hundred milliseconds later. The late
positive component, or LPC, is a positive going wave beginning
at approximately 600 ms after stimulus onset. The component is
modulated in some experimental designs by the same lexicalsemantic factors that modulate the N400 (Van Petten & Luka,
2006). In others it is modulated by repetition and has been linked
to explicit memory processes (e.g., Olichney et al., 2000). Using
the current experimental design, we previously observed a modulation of the LPC effect over anterior electrodes in response
to word-onset ortho-phonological overlap, which we referred
to as a post-lexical anterior negativity or PLAN (Justus et al.,
2009). Other authors have reported similar reverse-direction repetition effects, including Holcomb, Anderson, and Grainger (2005,
Exp. 2), who observed it in a study of repetition priming, and
Van Petten, Macizo, and O’Rourke (2007), who observed it in
a study of final-syllable-overlap ortho-phonological priming. We
suggested that PLAN modulation of the LPC reflects a relatively
late, strategic process driven by word-onset ortho-phonological
overlap.
Unlike the N400, there is evidence that neural generators of
the LPC include the inferior frontal regions targeted in the current
project. Using Kutas and Hillyard (1980) sentence–context design,
Swick et al. (1998) showed that patients with damage to inferior
frontal regions failed to demonstrate modulation of the LPC by
semantic context, suggesting that the main differences between
frontal patients and controls lay not in lexical-semantic access,
but rather in post-lexical processes beginning around 600 ms following the onset of the critical word. To our knowledge, neither
the dual-system theory nor the single-system theory currently
acknowledges or makes predictions regarding post-lexical processes in the past-tense priming task, including the contribution
of strategies driven by ortho-phonological overlap.
The inclusion of ERP data in the current study as well as other
recent work employing immediate morphological priming (Justus
et al., 2008, 2009; Kielar & Joanisse, 2010) complements a larger
literature concerning the electrophysiology of morphological processing, which has primarily employed either the delayed-priming
design (Münte, Say, Clahsen, Schiltz, & Kutas, 1999; RodríguezFornells, Münte, & Clahsen, 2002; Weyerts, Münte, Smid, & Heinze,
1996) or an incorrect-inflection design (Gross, Say, Kleingers,
Clahsen, & Münte, 1998; Morris & Holcomb, 2005; Newman,
Ullman, Pancheva, Waligura, & Neville, 2007; Penke et al., 1997;
Rodríguez-Fornells, Clahsen, Lleó, Zaake, & Münte, 2001; Weyerts,
Penke, Dohrn, Clahsen, & Münte, 1997). The current work, in which
neuropsychological and ERP methods are combined, also complements the literature concerning the N400 in aphasia (Friederici,
Hahne, & von Cramon, 1998; Friederici, von Cramon, & Kotz, 1999;
Hagoort, Brown, & Swaab, 1996; Swaab, Brown, & Hagoort, 1997,
1998).
2. Methods
2.1. Participants
To maximize the neural homogeneity of the patient group, we adopted lesionbased rather than symptom-based inclusion criteria for this study. Eleven patients
were recruited based on damage to Broca’s area, operationally defined as the pars
5
Fig. 1. Brain lesions of each of the eleven patients with damage to Broca’s area, defined as the pars opercularis (BA 44) and pars triangularis (BA 45) of the LIFG, are shown in
red on a standard brain image. An overlay of the eleven brain lesions illustrates the regions of maximum overlap in the LIFG and insula. These images were produced using
MRIcro (Rorden & Brett, 2000).
opercularis (BA 44) and pars triangularis (BA 45) of the LIFG, following a single
infarction in the precentral branch of the middle cerebral artery. Lesions were documented by CT or MRI scans and reconstructed onto a standard brain template of
the Montréal Neurological Institute using MRIcro software (Rorden & Brett, 2000).
These lesion reconstructions are presented for every patient in Fig. 1, and the extent
of damage in the pars opercularis, pars triangularis, pars orbitalis, and the insula is
provided in Table 1.
All patients had a history of clinically significant aphasia. However, no specific
aphasia diagnosis was required for inclusion in this study. The patients typed as
four anomic aphasics, one Broca’s aphasic, three unclassifiable aphasics, and three
within normal limits (WNL) on the most recent administration of the Western Aphasia Battery (WAB). The subscores of the WAB are also provided in Table 1. Note that,
had we adopted the requirement that all patients have a diagnosis of Broca’s aphasia, the size and anatomical diversity of the patients’ lesions would have increased
dramatically, as a chronic Broca’s aphasia requires a large lesion encompassing the
frontal and anterior temporal cortices, underlying white matter, the insula, and basal
ganglia, and which does not necessarily include Broca’s area in the LIFG (Dronkers,
Plaisant, Iba-Zizen, & Cabanis, 2007; Dronkers & Baldo, 2009). Therefore, a lesionbased inclusion criterion was the most appropriate for testing the hypothesis that
the absence of significant regular-verb priming is specifically associated with damage to the LIFG (Tyler, de Mornay Davies et al., 2002; Tyler, Marslen-Wilson, et al.,
2005).
The eleven patients (aged 58 ± 12 years; education 16 ± 3 years; 4 women, 7
men) were matched to eleven controls (aged 57 ± 12 years; education 16 ± 3 years;
3 women, 8 men). All participants were right-handed and native speakers of US
English. All procedures were approved by the institutional review board of the VA
Northern California Health Care System, and written informed consent was obtained
from all participants.
Table 1
Participants.
Patient
Sex
Age
Ed.
Years
post
% Pars
oper.
P1
P2
P3
P4
P5
P6
P7
P8
P9
P10
P11
Patient group
F
F
M
F
M
M
M
M
M
F
M
4F, 7M
54
63
68
69
36
42
65
65
72
47
54
58 ± 12
17
18
20
15
13
18
12
16
12
20
16
16 ± 3
4
6
2
4
6
3
16
12
8
7
8
7±4
7
91
79
98
99
97
100
86
100
100
100
87
Control group
3F, 8M
57 ± 12
16 ± 3
% Pars.
trian.
47
50
41
64
93
95
100
60
49
100
98
72
% Pars
orbit.
% Insula
WAB
fluency
WAB
comp.
WAB
rep.
WAB
naming
WAB
type
9
6
5
33
73
10
81
96
2
97
56
43
19
58
71
73
99
64
99
100
53
90
74
73
10
9
10
9
2
9
6
4
9
9
5
7
9.80
8.85
10.00
9.75
9.45
8.70
8.90
10.00
9.95
10.00
9.20
9.51
10.0
8.8
9.8
9.4
5.4
8.9
9.4
5.2
9.4
8.6
7.5
8.4
9.5
8.9
9.5
8.8
8.9
9.3
6.3
6.5
9.0
9.2
8.7
8.6
WNL
Anomic
WNL
Anomic
Unclass.
Unclass.
Anomic
Broca
WNL
Unclass.
Anomic
4 anomic, 1 Broca,
3 unclass., 3 WNL
6
Table 2
Auditory immediate-priming design with lexical decision.
Target
Target freq.a
Target length
Look
Seem
Tarb
1.94 ± 0.4
1.94 ± 0.4
528 ± 92 ms
523 ± 78 ms
Spoke
Bound
Took
Speak
Wake
*
Plinn
1.98 ± 0.8
1.98 ± 0.8
515 ± 92 ms
518 ± 81 ms
Primed
Unprimed
Nonword
Bead
Bulb
Deer
Bee
Pie
*
Cleeth
1.61 ± 1.1
1.59 ± 1.1
516 ± 53 ms
522 ± 65 ms
Primed
Unprimed
Nonword
Barge
Bribe
Pouch
Bar
Tea
Gwal
1.29 ± 0.9
1.21 ± 0.8
494 ± 70 ms
496 ± 64 ms
Word Type
Priming
Prime
Regular
verbs
(n = 150)
Primed
Unprimed
Nonword
Looked
Worked
Asked
Irregular
verbs
(n = 150)
Primed
Unprimed
Nonword
Pseudopast
(n = 150)
Orthophono
(n = 150)
a
*
*
Log lemma frequency per million, ±s.d.
2.2. Procedure
Following the informed consent procedure and application of the recording
electrodes, participants were seated in a sound-attenuated booth. Each trial of the
experiment began with fixation for 1500 ms, followed by the prime, spoken by a
female voice. The target was presented 1400 ms after the onset of the prime (an
average of 870 ms ISI, given mean prime length of 530 ms), spoken by a male voice.
Participants were instructed to identify quickly and accurately whether the target (male) voice was saying a real word or a nonword. Given that seven of the
patients demonstrated some degree of right-sided hemiparesis, all participants were
required to respond using the left hand, and were encouraged to use the middle and
index fingers to respond to nonwords and real words, respectively.
designed by choosing items of the same word class, frequency, and, to the extent
possible, number of phonemes and letters as in the primed condition.
Orthophono: 50 pairs of semantically unrelated words were selected in which
the prime word differed from the target word by the addition of a single phoneme
and one or two letters (e.g., barge–bar). Unlike in the pseudopast condition, this
addition was not phonologically consistent with the inflectional rhyme pattern of
the regular past tense, and the spelling of the shorter target word was completely
contained within the spelling of the longer prime word. These 50 orthophono primetarget pairs were all used in the primed condition. As in the pseudopast condition,
the items in the unprimed and nonword conditions were designed by choosing items
of the same word class, frequency, and number of phonemes and letters as in the
primed condition.
The entire stimulus set may be viewed in Appendix of Justus et al. (2008).
2.3. Design and selection of stimuli
2.4. Stimulus recording
The 1200 stimulus items of Justus et al. (2008) were used, resulting in 600
prime-target pairs according to a 4 (regular, irregular, pseudopast, orthophono) by
3 (primed, unprimed, nonword) design with 50 trials per cell (Table 2). All cells consisted of 46 monosyllabic prime-target pairs and 4 bisyllabic prime-target pairs. All
of the 600 primes and 400 of the 600 targets were real words, selected with the aid of
the CELEX Lexical Database (Baayen, Piepenbrock, & Gulikers, 1995). The remaining
200 targets were selected using the ARC Nonword Database (Rastle, Harrington, &
Coltheart, 2002).
One design constraint that was adopted given the use of ERP data (requiring at
least 50 items per cell), the desire to avoid repetition effects on the N400 (observed
even for extremely long lags; Bentin & Peled, 1990; Nagy & Rugg, 1989), and the
small number of available stimuli in some conditions (especially the pseudopast
condition) was the decision to use a between-item design. Accordingly, extreme
care was taken to balance the items used in the primed and unprimed conditions
– as well as the prime words in the nonword condition – on factors such as lemma
frequency, syllabicity, word class, number of phonemes, number of letters, and in
the case of irregular verbs, the type of irregularity. This was done simultaneously
with the constraint to balance the same factors as closely as possible between the
regular, irregular, pseudopast, and orthophono conditions. Further constraints on
the selection of stimulus items for each Word Type were as follows.
Regular verbs: 150 regular verbs were selected and divided among the primed,
unprimed, and nonword conditions. All were the /t/ or /d/ rather than the /Id/ allomorph, in order to control for syllabicity with the other conditions. Regular past
tenses that shared pronunciations with other words were avoided (e.g., packed/pact
or missed/mist).
Irregular verbs: 150 of the ∼180 irregular English verbs were selected. Modal
forms (e.g., could–can, would–will) were avoided, as were words that are typically regularized in US English (e.g., learnt–learn, spilt–spill). The chosen items were
divided among primed, unprimed, and nonword conditions as follows. First, to avoid
repetition priming, the 26 no-change irregulars (e.g., put) were used as primes
only in the nonword condition. The remaining 45 weak irregulars in the set (e.g.,
spent–spend) were divided between the primed (n = 23) and unprimed (n = 22) conditions. The rest of the words in the three conditions (n = 27, 28, and 24) were strong
verbs (e.g., spoke–speak), or, in two cases, suppletive verbs (was–is, went–go). Care
was taken to distribute the subordinate families of irregular verbs as evenly as
possible throughout the three conditions, given these constraints.
Pseudopast: 50 pairs of semantically unrelated words were selected in which
the prime word differed from the target word by the addition of /t/ or /d/, in a
manner phonologically consistent with the inflectional rhyme pattern of the regular past tense (e.g., bead–bee). Potential stimuli overlapping with items in the verb
conditions (e.g., field–feel, bide–buy) or sharing pronunciations with regular verbs
of a higher word frequency were avoided. Given the small number of appropriate monosyllabic pairs available in English, these 50 pseudopasts were all used in
the primed condition. The items in the unprimed and nonword conditions were
Sound files were digitally recorded in a sound-attenuated booth by two native
speakers of US English, one woman (prime words) and one man (target words),
who were naïve to the purpose of the experiment. The speakers were coached in
pronouncing the real words and nonwords correctly, and in delivering all items
clearly and with a consistent intonation, sound level, and speed. A pseudorandom
recording order was created to ensure that no changes in speech over the course of
the recording session would correlate with experimental factors.
Recordings were later filtered of white noise, edited into individual sound files,
and further normalized for sound level. Analyses of sound file lengths confirmed
that these were properly balanced (see Table 2).
2.5. EEG recording
Electroencephalographic (EEG) activity was recorded using 26 electrodes
embedded in an electrode cap. These electrodes consisted of the 19 electrodes
of the International 10–20 system (Jasper, 1958) plus 7 additional electrodes at
positions AF3/4, FC5/6, CP5/6, and POz. Four external electrodes recorded the left
mastoid, right mastoid, left-horizontal electrooculogram (EOG), and left-vertical
EOG. During acquisition, EEG was recorded with reference to the left mastoid. Signals were amplified (×20,000), filtered (0.1–80 Hz), and digitized at a sampling
rate of 256 Hz (SA Instrumentation). Data were later re-referenced to the average
of the two mastoids and digitally low-pass filtered (20 Hz). Eye-blinks that were
uncontaminated by additional artifacts were corrected using an adaptive filtering
algorithm.
Mean amplitudes were calculated relative to a 100-ms window preceding
word onset. No main effects of Electrode are reported. Interactions involving Electrode are reported with uncorrected F values and degrees of freedom, and with
Greenhouse–Geisser corrected p values.
Topographic maps were created by calculating voltage differences between the
unprimed and primed conditions at each electrode and interpolating voltage differences for the rest of the scalp using a spherical spline mapping method (Perrin,
Pernier, Bertrand, & Echallier, 1989).
3. Results
3.1. Behavioral data
The first analyses of the behavioral data address the first two of
the four main questions posed by this paper, namely whether the
priming of regular and irregular verbs dissociates following damage to Broca’s area, and whether this dissociation can be explained
7
Fig. 2. Response time data, first as a function of Word Type (upper four plots) and second for the irregular verbs alone, separated into weak irregular verbs and strong verbs
(lower four plots). Controls are presented to the left and patients to the right. Bar graphs illustrate the grand mean response times for each group, with error bars representing
standard error. Scatter plots illustrate individual response time data, using a difference score in order to represent each person by a single point. Horizontal bars represent
the mean difference for the group.
8
in terms of ortho-phonological priming (Fig. 2, upper four plots).
Behavioral data were first analyzed using four analyses of variance with the factors Word Type (regular, irregular, pseudopast,
orthophono) and Priming (primed, unprimed), separating the controls and the patients, for the response time and error rate data.
These were followed by additional analyses comparing the degree
of Priming for each possible pair of Word Types. A final group of
analyses evaluated whether the degree of Priming for each Word
Type differed significantly between controls and patients. For the
response time analyses, only correct responses to real words were
used. Such responses falling beyond the individual mean plus three
standard deviations were removed (a mean of 2.2 instances per
person for the patients and 3.2 for the controls).
The main effect of Priming was reliable for the controls in
both response times and errors [RT: F(1,10) = 28.4, p < 0.001; errors:
F(1,10) = 7.5, p = 0.02] but for the patients was reliable only in the
errors [RT: F(1,10) = 1.7, p = 0.23; errors: F(1,10) = 16.7, p = 0.002].
The main effect of Word Type was reliable in both measures for
both the controls [RT: F(3,30) = 35.8, p < 0.001; errors: F(3,30) = 34.4,
p < 0.001] and the patients [RT: F(3,30) = 28.7, p < 0.001; errors:
F(3,30) = 16.5, p < 0.001]. Finally, the interaction between Priming
and Word Type was reliable in response times for the controls
[RT: F(3,30) = 7.3, p = 0.001; errors: F(3,30) = 1.8, p = 0.16] and in
both response times and errors for the patients [RT: F(3,30) = 10.0,
p < 0.001; errors: F(3,30) = 40.6, p = 0.02].1
For the controls, significant RT priming was found for regular verbs [RT: 44 ms, t(10) = 4.0, p = 0.002; errors: 1.1%, t(10) = 2.2,
p = 0.05], irregular verbs [RT: 102 ms, t(10) = 6.0, p < 0.001; errors:
0.7%, t(10) = 1.2, p = 0.27], pseudopast [RT: 40 ms, t(10) = 2.5,
p = 0.03; errors: 2.9%, t(10) = 3.0, p = 0.02], and orthophono [RT:
32 ms, t(10) = 2.3, p = 0.04; errors: 1.6%, t(10) = 1.5, p = 0.17]. Further, greater RT priming was observed for irregular verbs compared
to regular verbs [F(1,10) = 29.6, p < 0.001], irregular verbs compared to pseudopast [F(1,10) = 9.7, p = 0.01], and irregular verbs
compared to orthophono [F(1,10) = 17.4, p = 0.002]. Greater error
priming was observed for the pseudopast compared to irregular
verbs [F(1,10) = 5.7, p = 0.04].
For the patients, significant RT priming was found only
for the irregular verbs [98 ms, t(10) = 5.3, p < 0.001; errors:
2.5%, t(10) = 2.1, p = 0.06], and not for regular verbs [RT: 5 ms,
t(10) = 0.2, p = 0.82; errors: 1.8%, t(10) = 2.3, p = 0.09], pseudopast [RT: −23 ms, t(10) = −0.8, p = 0.43; errors: 8.2%, t(10) = 3.5,
p = 0.006], or orthophono [RT: −2 ms, t(10) = 0.1, p = 0.89; errors:
3.6%, t(10) = 2.3, p = 0.04]. However, error effects or trends were
observed for all four Word Types. The lack of a significant regular
priming effect is not dependent on the inclusion of P7 (see Fig. 2);
if P7 were removed from the analysis, the regular RT effect would
increase to a trend but remain nonsignificant (20 ms, t(9) = 1.5,
p = 0.17). Further, greater RT priming was observed for irregular
verbs compared to regular verbs [F(1,10) = 20.5, p = 0.001], irregular verbs compared to pseudopast [F(1,10) = 13.3, p = 0.004], and
irregular verbs compared to orthophono [F(1,10) = 18.0, p = 0.002].
Greater error priming was observed for pseudopast compared to
irregular verbs [F(1,10) = 6.2, p = 0.03] and pseudopast compared
to regular verbs [F(1,10) = 8.3, p = 0.02].
To ascertain whether priming for each Word Type was
significantly different between the controls and the patients,
the interaction between Priming and Group was examined for
each Word Type separately. The interaction did not reach sig-
nificance for regular verbs [RT: F(1,20) = 2.9, p = 0.10; errors:
F(1,20) = 0.5, p = 0.51], irregular verbs [RT: F(1,20) = 0.04, p = 0.85;
errors: F(1,20) = 1.8, p = 0.20], pseudopast [RT: F(1,20) = 3.8, p = 0.07;
errors: F(1,20) = 4.3, p = 0.05], or orthophono [RT: F(1,20) = 2.9,
p = 0.11; errors: F(1,20) = 1.1, p = 0.31]. However, the trends support
the results of the separate analyses that priming was reduced for
the patients relative to the controls for regular verbs, pseudopast,
and orthophono, and preserved for irregulars.
Correlations within the response time data: Given that a major
interest of this study was to explore the relationship between
regular-verb priming and word-onset ortho-phonological priming,
one might ask whether the size of the regular priming effect covaried with that of the pseudopast and orthophono conditions. The
relationship between priming effect sizes and mean response time
was also of interest. To address this, a series of Pearson correlations
was performed on the difference between primed and unprimed RT
for each Word Type, along with the mean response time (5 items,
10 pairings), using the Holm–Bonferroni correction. Mean response
time (specifically, mean correct response time to all real words) was
negatively correlated with the size of the regular [r(20) = −0.68,
p < 0.001], pseudopast [r(20) = −0.76, p < 0.001], and orthophono
[r(20) = −0.62, p = 0.002] priming effects, but not with the irregular
priming effect [r(20) = −0.12, p = 0.60]. Further, the regular priming
effects correlated with the pseudopast [r(20) = 0.55, p = 0.008] and
orthophono effects [r(20) = 0.55, p = 0.009], but irregular priming
effects did not [each p > 0.35]. Finally, regular and irregular priming effects correlated [r(20) = 0.52, p = 0.01], as did pseudopast and
orthophono effects [r(20) = 0.48, p = 0.02]; however, these last correlations do not survive correction for multiple comparisons.
Effects of the type of irregularity: To address the third main question posed by this paper, namely whether the regular–irregular
dissociation observed in the patient group would be better
described as categorical or continuous, the data for the irregular
verbs were divided into two subsets – weak irregular verbs and
strong verbs (Fig. 2, lower four plots). Two analyses with the factors Irregular Subset (weak irregular, strong) and Priming (primed,
unprimed) tested for an interaction between these variables, followed by sets of two planned comparisons to test for priming
in each irregular subset alone. The interaction was significant in
the RT data for both the controls [RT: F(1,10) = 32.3, p < 0.001;
errors: p = 0.55] and the patients [RT: F(1,10) = 11.7, p = 0.007;
errors: p = 0.14], such that greater priming was observed for the
strong verbs compared to the weak irregular verbs. Both the weak
irregular verbs and the strong verbs resulted in significant RT priming when examined separately, for both the controls [weak irregular RT: 58 ms, t(10) = 2.6, p = 0.03; errors: p = 0.19; strong RT: 162 ms,
t(10) = 8.2, p < 0.001; errors: p = 0.72] and the patients [weak irregular RT: 74 ms, t(10) = 3.8, p = 0.003; errors: p = 0.59; strong RT:
154 ms, t(10) = 6.1, p < 0.001; errors: t(10) = 2.3, p = 0.05], with the
patient strong-verb effect also reaching significance for the errors.
To complete the evaluation of the three verb classes, comparisons were made between the regular verbs and each subset of the
irregular verbs. Unsurprisingly given the results above, strong verbs
resulted in greater RT priming than did the regular verbs, for both
the controls [RT: F(1,10) = 50.0, p < 0.001; errors: p = 0.52] and the
patients [RT: F(1,10) = 37.3, p < 0.001; errors: p = 0.19]. However, the
weak irregular verbs primed more strongly than did the regular
verbs for just the patients [RT: F(1,10) = 8.5, p = 0.02; errors: p = 0.28]
but not for the controls [RT: F(1,10) = 0.9, p = 0.37; errors: p = 0.83].
3.2. ERP data
1
The direction of significant main effects of Priming and Word Type were consistent across RTs and errors (i.e., conditions associated with faster RTs had fewer
errors). Priming by Word Type interactions tended to be reversed for RTs and errors
(i.e., Word Types associated with larger RT priming effects had smaller error priming
effects).
The remainder of the analysis addresses the fourth main
question, concerning whether the regular–irregular dissociation
extends to event-related potentials. Visual inspection of the
averaged waveforms revealed the N400 component, which was
centered over central-posterior electrodes. In both groups and for
all four word types, the N400 component was numerically less negative for primed trials compared to unprimed trials. Accordingly,
a nine-electrode region of interest (Cz, Pz, POz, C3, C4, CP5, CP6,
P3, P4) was chosen to examine the relative sizes of the ERP effects
across the four conditions (Fig. 3). Following the N400, during the
late positive component of the waveform, two effects emerged:
in some cases the N400 priming effect was preserved through the
LPC (see especially the irregular verbs), whereas for other conditions the waveforms crossed at approximately 600 ms, such that
the LPC was more negative for primed trials than for unprimed trials (see especially the pseudopast). Given these observations, two
temporal windows were chosen for the analysis – an early window of 350–600 ms, to examine the N400 component, and a late
window of 600–850 ms, to examine the LPC (Fig. 4). These analysis windows were slightly delayed and elongated relative to those
used in our previous intra-modal study of healthy young controls
(300–500 and 500–700 ms), consistent with previous work concerning the N400 component in normal aging (Federmeier & Kutas,
2005) and in aphasia (Hagoort et al., 1996; Swaab et al., 1997).
9
ERP data were first analyzed using four analyses of variance with
the factors Word Type (regular, irregular, pseudopast, orthophono),
Priming (primed, unprimed), and Electrode (9-electrode ROI),
separating the controls and the patients, as well as the N400
(350–600 ms) and LPC (600–850 ms) windows. These were followed by additional follow-up analyses to compare the relative
sizes of the various priming effects, and whether they differed significantly between controls and patients, in a manner parallel to
the response time and error analyses.
N400 window (350–600 ms): For the controls during the N400
window, main effects of Priming [F(1,10) = 9.0, p = 0.01] and Word
Type [F(3,30) = 4.2, p = 0.01] were reliable, as well as an interaction trend between these variables [F(3,30) = 2.4, p = 0.08]. When
considering the four Word Types individually, reliable N400
priming effects were observed for regular verbs [F(1,10) = 4.9,
p = 0.05] and irregular verbs [F(1,10) = 29.9, p < 0.001]. A marginal
effect was observed for the pseudopast condition [F(1,10) = 4.2,
p = 0.07], whereas no effect was observed for the orthophono
condition [F(1,10) = 0.6, p = 0.46]. The N400 reduction for primed
irregular verbs was significantly larger than that for pseu-
Fig. 3. ERPs to target words within the nine-electrode region of interest, demonstrating the N400 effect as a function of Word Type (regular verbs, irregular verbs, pseudopast,
and orthophono) and Priming. Controls are presented to the left and patients to the right.
10
Fig. 4. Topographic maps of the scalp distributions displaying the N400 priming effect (unprimed-primed) between 350–600 ms and 600–850 ms as a function of Word Type
(regular verbs, irregular verbs, pseudopast, and orthophono) and Priming. Controls are presented to the left and patients to the right.
dopast [F(1,10) = 6.0, p = 0.03] and marginally larger than that for
orthophono [F(1,10) = 3.8, p = 0.08]. The other four pairings were
not significant (p > 0.20).
For the patients during the N400 window, main effects of
Priming [F(1,10) = 23.1, p = 0.001] and Word Type [F(3,30) = 3.1,
p = 0.04] were reliable. The interaction between Priming and
Word Type was not (p = 0.99). When considering the four Word
Types individually, reliable or marginal N400 priming effects were
observed for both regular verbs [F(1,10) = 8.8, p = 0.01] and irregular verbs [F(1,10) = 3.7, p = 0.08], and further, for the pseudopast
[F(1,10) = 10.2, p = 0.01] and orthophono [F(1,10) = 18.5, p = 0.002]
conditions. None of these N400 reductions were significantly larger
than the others (six pairings, each p > 0.70).
To determine whether the N400 priming effect for each Word
Type was significantly different between the controls and the
patients, the interaction between Priming and Group was examined for each Word Type separately. The interaction was not
significant for regular verbs [F(1,20) = 0.4, p = 0.51], irregular verbs
[F(1,20) = 0.001, p = 0.98], pseudopast [F(1,20) = 2.1, p = 0.16], or
orthophono [F(1,20) = 1.9, p = 0.19].
LPC window (600–850 ms): For the controls during the LPC
window, the main effect of Word Type [F(3,30) = 8.1, p < 0.001]
and the interaction between Priming and Word Type were reliable [F(3,30) = 5.7, p = 0.003]. The main effect of priming was not
(p = 0.80). When considering the four Word Types individually, significant priming was observed for irregular verbs [F(1,10) = 6.9,
p = 0.03], in the same direction as the N400 priming effect. A
marginal reversed effect was observed for the pseudopast condition [F(1,10) = 4.6, p = 0.06], such that primed items were more
negative than unprimed items. Neither the regular (p = 0.72) nor
the orthophono (p = 0.27) effect was significant during this later
window. Considering the Word Types in pairs, the irregular effect
was significantly different from both the pseudopast [F(1,10) = 14.2,
p = 0.004] and orthophono [F(1,10) = 11.3, p = 0.007] effects. The
regular effect trended towards a difference from pseudopast
[F(1,10) = 3.2, p = 0.10] and orthophono [F(1,10) = 3.9, p = 0.08]. The
other two pairings were not significant (p > 0.18).
For the patients during the LPC window, main effects of Priming [F(1,10) = 7.8, p = 0.02] and Word Type [F(3,30) = 5.0, p = 0.006]
were reliable. The interaction between Priming and Word Type was
not (p = 0.17). When considering the four Word Types individually,
significant priming was observed for irregular verbs [F(1,10) = 7.6,
p = 0.02] and orthophono [F(1,10) = 5.8, p = 0.04], both in the same
direction as the N400 priming effect. The effect for regular verbs
was not significant during this later window [F(1,10) = 1.5, p = 0.25],
nor was the effect for pseudopast [F(1,10) = 0.9, p = 0.37], despite
this condition showing a numerically reversed effect as in the
controls. Consistent with this observation, the pseudopast effect
trended towards a difference from both the irregular [F(1,10) = 4.3,
p = 0.07] and orthophono [F(1,10) = 3.5, p = 0.09] effects. The other
four pairings were not significant (p > 0.20).
To determine whether the LPC priming effect for each word type
was significantly different between the controls and the patients,
the interaction between Priming and Group was examined for
each Word Type separately. The interaction was not significant for
regular verbs [F(1,20) = 0.6, p = 0.44], irregular verbs [F(1,20) = 0.5,
p = 0.47], or pseudopast [F(1,20) = 0.4, p = 0.51], and just reached
significance for orthophono [F(1,20) = 4.4, p = 0.05].
Effects of the type of irregularity: Potential effects of the type
of irregularity within the irregular word set on the N400 and LPC
were explored using analyses of variance with the factors Irregular
Subset (weak irregular, strong), Priming (primed, unprimed), and
Electrode (9-electrode ROI), specifically to examine for an interaction between Irregular Subset and Priming. Unlike in the behavioral
data, this interaction was not significant during the N400 window
for either the controls (p = 0.12) or the patients (p = 0.61), nor was it
significant during the LPC window for either the controls (p = 0.72)
or the patients (p = 0.93).
Correlations within the ERP data: As before, one might ask
whether the size of the N400 priming effect for regular verbs covaried with that of the pseudopast and orthophono conditions.
A series of Pearson correlations was performed on the average
difference in the nine-electrode ROI for each Word Type during the N400 and LPC windows (8 items, 28 pairings), using the
Holm–Bonferroni correction. The N400 priming effects for the
pseudopast and orthophono conditions correlated [r(20) = 0.64,
p = 0.002], but these effects did not correlate with the N400 priming
effect for regular verbs (each p > 0.40). Further, the N400 priming
effect for each Word Type correlated with the corresponding LPC
effect (each p < 0.002).
Correlations between the response time and ERP data: Given the
relationships found between regular, pseudopast, and orthophono
priming effects in the behavioral data, and the lack of such relationships in the N400 data, one might ask whether the size of the
behavioral priming effects predicted the size of the N400 priming
effects, either for the corresponding condition or between conditions. A series of Pearson correlations was performed to compare
the mean response time to all real words, the four RT difference
scores for each Word Type, and the eight ERP difference scores
(40 pairings). None of these correlations were significant after the
Holm–Bonferroni correction was applied.
3.3. Lesion and Western Aphasia Battery data
A final analysis was carried out to determine whether the lesion
extent in our region of interest, LIFG pars opercularis (BA 44) and
pars triangularis (BA 45), predicted the size of the behavioral or ERP
priming effects. To explore the lesion variability outside of Broca’s
area, ten other brain regions that were implicated in some members of the current patient group were included in this analysis:
LIFG pars orbitalis, insula, middle frontal gyrus, rolandic operculum,
putamen, precentral gyrus, postcentral gyrus, superior temporal
pole, superior temporal gyrus, and Heschl’s gyrus. Similarly, the
four subscores of the Western Aphasia Battery (fluency, comprehension, repetition, and naming), which are used to determine WAB
aphasia type, were entered into this analysis. Only one correla-
11
tion survived correction for multiple comparisons: lesions affecting
a greater percentage of the insula were associated with smaller
ERP priming effects for the irregular verbs in the N400 window
[r(9) = 0.93, p < 0.001]. None of the correlations between a behavioral priming score and either lesion or WAB data approached
significance, even when uncorrected for multiple comparisons.
Note that because the patients were selected for a relatively high
degree of lesion uniformity (i.e., all affecting the pars opercularis
and pars triangularis), as opposed to the unselected patient population that might be used in a true lesion-symptom mapping analysis
(e.g., Bates et al., 2003; Tyler, Marslen-Wilson, et al., 2005), correlations implicating these two regions were not expected.
4. Discussion
The present study first confirmed previous observations (Justus
et al., 2008, 2009; Longworth, Keenan, et al., 2005; Marslen-Wilson
& Tyler, 1997; Tyler, de Mornay Davies, et al., 2002; Tyler et al.,
2004) that, in neurologically healthy people, hearing the pasttense form of both regular and irregular English verbs facilitated
an auditory lexical decision to the corresponding present-tense
form (e.g., looked–look, spoke–speak). This behavioral facilitation
was accompanied by a reduction in the N400 component to the target word. Here we tested eleven patients with damage to Broca’s
area, i.e., the LIFG pars opercularis (BA 44) and pars triangularis
(BA 45), along with eleven matched controls, to see whether a
previously observed behavioral dissociation between regular and
irregular morphological priming (Marslen-Wilson & Tyler, 1997;
Tyler, de Mornay Davies, et al., 2002) would replicate in this welldefined patient group. Second, we asked whether the absence
of regular-verb priming could be explained by the phonological
relationship between prime and target, by looking for the corresponding absence of two phonological priming effects, one in which
prime and target were related by the inflectional rhyme pattern
of the regular past tense, and another in which prime and target were related by other kinds of phonological and orthographic
overlap. Third, we asked whether such regular–irregular dissociations would be better described as categorical or graded. Finally,
we asked whether the above effects could be captured with eventrelated potentials, with a focus on the N400 and LPC components
to address whether lexical or post-lexical processes were disrupted
in the patient group. We consider each of these in turn.
4.1. Does the priming of regular and irregular verbs dissociate
following damage to Broca’s area?
Regarding the first major question posed by this paper, the previously observed dissociation between regular- and irregular-verb
priming (Marslen-Wilson & Tyler, 1997; Tyler, de Mornay Davies,
et al., 2002) was replicated in the response time data of the current study. As a group, controls demonstrated significant priming
of 44 ms for regular verbs (p = 0.002), as well as significant – and significantly greater – priming of 102 ms for irregular verbs (p < 0.001).
No differences in the error rates for regular and irregular verbs
were observed (1.1% and 0.7%, respectively). As a group, patients
with damage to Broca’s area demonstrated a non-significant priming effect of 5 ms for regular verbs (p = 0.82) and a significant – and
significantly greater – priming effect of 98 ms for irregular verbs
(p < 0.001). As before, no differences in the error rates for regular
and irregular verbs were observed (1.8% and 2.5%, respectively). We
note that in a direct comparison of controls and patients, there was
only a trend for the regular-verb priming effect to be smaller for
the patients compared to the controls (p = 0.10).
These findings permit a stronger connection to be drawn
between damage to Broca’s area and the disruption of regular-
12
verb priming than was possible in the studies of Marslen-Wilson
and Tyler (1997) and Tyler, de Mornay Davies, et al. (2002). The
patients in those studies had extremely diverse left-hemisphere
lesions that extended well beyond the LIFG. The eleven patients
included here were selected on the basis of damage to Broca’s area,
defined as the pars opercularis (BA 44) and pars triangularis (BA
45) of the LIFG. The present study also complements the finding
of a correlation between tissue density in the LIFG and the size of
regular-verb priming effects that was found in a study of 22 brainlesioned patients who were unselected for lesion location (Tyler,
Marslen-Wilson, et al., 2005).
Despite the relative lesion homogeneity of the current patient
group, not all patients were equivalent with regard to the regularverb priming effect (Fig. 2). Some of the LIFG patients demonstrated
preserved regular-verb priming effects of as much as 99 ms (patient
5). However, other patients demonstrated either flat effects (as in
Tyler, de Mornay Davies, et al., 2002) or inhibitory effects (as in
Marslen-Wilson & Tyler, 1997) of as much as 155 ms (patient 7),
thus reducing the mean effect for the group. As with the controls,
the difference between the priming effects for regulars and irregulars in the patient group was fairly consistent, with everyone except
for Patient 6 showing a numerically greater priming effect for irregulars compared to regulars. Neither the lesion extent outside of
Broca’s area nor the aphasia subscores on the Western Aphasia Battery provides a clear explanation for this variation in regular-verb
priming or any of the behavioral effects.
At face value, the lack of significant regular past-tense priming is
consistent with the dual-system perspective (e.g., Marslen-Wilson
& Tyler, 1998; Pinker, 1999; Pinker & Ullman, 2002), including its
IRP variant (Marslen-Wilson & Tyler, 2007), that the perception
and lexical access of regular past tenses involves an affix-stripping
process that is dependent on Broca’s area, whereas irregular past
tenses can bypass this process and be retrieved directly from the
mental lexicon. However, the single-system perspective also offers
explanations for this dissociation, with specific predictions about
the relationships between morphological and ortho-phonological
priming, as well as a prediction that the dissociation should be
graded and not categorical (e.g., Bird et al., 2003; Joanisse &
Seidenberg, 1999; McClelland & Patterson, 2002).
4.2. Is the regular–irregular dissociation related to phonological
factors, and if so, does the inflectional rhyme pattern have special
status?
Two control conditions were included in the experiment to
examine word-onset ortho-phonological priming. The pseudopast
condition (bead–bee) was designed to mimic the inflectional rhyme
pattern between regular past- and present-tense forms, which
is argued to hold a special morpho-phonological status in the
IRP viewpoint of Marslen-Wilson and Tyler (2007). This approach
predicts that regular-verb priming effects should pattern with
pseudopast priming effects, given that only conditions in which
an inflectional rhyme pattern is present should trigger an attempt
at LIFG-dependent morphological segmentation. In contrast, the
orthophono condition (barge–bar), while representing a similar
degree of ortho-phonological overlap between prime and target,
did not signal a potential morphological relationship. The IRP
viewpoint predicts that this condition should not implicate LIFGdependent processing and as a result, orthophono priming effects
should be distinct in pattern from those of the regular verbs and
pseudopasts. However, single-system approaches that emphasize
a link between regular morphology and phonology (e.g., Bird et al.,
2003; Joanisse & Seidenberg, 1999; McClelland & Patterson, 2002),
might predict that all three priming effects – regular verbs, pseudopast, and orthophono – should group together.
The behavioral data straightforwardly support the latter pattern. Whereas the controls demonstrated significant priming
effects for all three of these conditions, the patients with damage
to Broca’s area did not demonstrate significant priming for any of
them. By examining individual variability in priming, it can further
be shown that the size of the regular, pseudopast, and orthophono
priming effects all correlate significantly with one another. These
data suggest that the reduced priming effect for regular verbs
following damage to Broca’s area is based in a reduction of the
phonological priming effects that we observe in this design. Contrary to the prediction of the IRP viewpoint, this relationship was
not specific to the condition that contained the regular inflectional
rhyme pattern (the pseudopast condition), but rather extended
to other kinds of phonological and orthographic overlap that did
not signal morphological relationships (the orthophono condition).
Instead, the relationship between regular morphology and phonology more generally is consistent with the argument that deficits in
regular morphology are based in deficits of phonological processing.
4.3. Is the regular–irregular dissociation better described as
categorical or continuous?
Patients with damage to Broca’s area demonstrated preserved
priming for irregular verbs in this experimental design. This result
is predicted both by dual-system viewpoints, including the IRP variant, and by single-system viewpoints, but for different reasons.
From a dual-system perspective, the preserved effect is the result
of the irregular past-tense forms having distinct representation in
the mental lexicon, thus bypassing the supposed impairment in
regular morphological processing. From a single-system perspective, the preserved irregular priming effect is the result of a stronger
dependence on semantic representations in linking the past- and
present-tense forms, and a weaker dependence on phonological
representations, which are supposedly compromised.
In considering the nature of this preserved irregular-verb priming, we divided the irregular verbs into two groups – weak
irregular verbs and strong verbs – and compared these two
classes with the regular verbs. In the case of the controls, the
regular–irregular difference is better described as a weak-strong
difference, with regular-verb (44 ms) and weak-irregular-verb
effects (58 ms) grouping together statistically and each being different from that of the strong verbs (162 ms). In the case of the
patients, the regular–irregular difference is better described as
a three-way split between regular verbs (5 ms), weak irregular
verbs (74 ms), and strong verbs (154 ms), all of which were significantly different from one another. These results suggest that the
regular–irregular dissociation is graded and continuous, not categorical (see Joanisse & Seidenberg, 1999, 2005; Justus et al., 2008,
2009; Kielar & Joanisse, 2010; Kielar et al., 2008; also see Hay &
Baayen, 2005).
These results are problematic for dual-system theories, which
argue for a categorical distinction between the processing of regular and irregular verbs. From this perspective, regular verbs are
processed according to a linguistic rule that adds or strips the suffix |-d|, realized as one of three different allomorphs (/d/, /t/, or /Id/),
to or from the present-tense stem. These rule-based morphological processes come into play only for the class of regular verbs, and
cannot be applied to the irregulars, which must be fully listed in
lexical memory. Even the weak irregular verbs – those that end in
/t/ or /d/ in the past tense, often as the result of affixation – are
not processed based on linguistic rules, but are memorized just as
the strong verbs are. Most dual-system perspectives (e.g., Pinker,
1999; Pinker & Ullman, 2002) therefore predict that any dissociations in the processing of regular and irregular verbs should be
categorical, with all regular verbs grouping together and all irreg-
ular verbs grouping together, regardless of whether they are weak
or strong.
The predictions of the IRP viewpoint (Marslen-Wilson & Tyler,
2007) are less clear. One might argue that this view predicts that
regular–irregular dissociations should be categorical, given that IRP
is clearly aligned with the dual-system framework of morphological decomposition for regular verbs only. However, one might also
argue that the IRP viewpoint can accommodate some continuous
dissociations. For instance, it is possible that the phonology of the
weak irregular past tenses could gain them access to the same LIFGdependent morphological segmentation mechanisms proposed for
regular verbs, resulting in an attempt at morphological segmentation (see Post et al., 2008). In the current study, about half of
the weak irregular past tenses contained the inflectional rhyme
pattern (e.g., slept, but not spent). However, this would be a more
plausible explanation for why such words might prime – or fail to
prime – the corresponding nonword that would result from IRPtriggered morphological decomposition (e.g., slept–slep), and not
the true present-tense form (e.g., slept–sleep).
The continuous regular–irregular dissociation that we report
here is particularly consistent with single-system theories, because
the result was specifically predicted. Single-system theories argue
that the processing of all verb morphology, both regular and irregular forms, entails the mapping between form (phonology and
orthography) and meaning (semantics), and that morphological
relations between words are emergent and reflect interactions
between these relationships (see Gonnerman, Seidenberg, &
Andersen, 2007). This perspective eschews the idea of discrete
linguistic rules, and instead embraces the notions of distributed
knowledge representation and probabilistic processing. The singlesystem perspective predicts that any dissociations in the processing
of regular and irregular verbs should be graded and continuous,
and rooted in differences in the ortho-phonological and semantic
relationships between present- and past-tense forms.
4.4. Does the regular–irregular dissociation extend to
event-related potentials?
Unlike in the behavioral data, the ERP priming effects for both
regular and irregular verbs were similar to each other for both
patients and controls (Fig. 3). Considering the overall comparison of
regular and irregular verbs, some subtle differences were observed:
for both controls and patients, the priming effect was significant
for both regulars and irregulars during an earlier 350–600 ms window targeting the N400, but either remained significant or trended
in this direction for the irregulars only during a later 600–850 ms
window targeting the LPC. Although the differential size of the
two effects did not reach significance, as in our study of healthy
young controls (Justus et al., 2008), the trend is consistent with
the argument of that study, as well as the current behavioral data,
that irregulars prime more strongly in this experimental design.
The more critical question for our purposes, however, concerns
whether the N400 verb-priming effect differed between controls
and patients, and whether this interacted with verb regularity.
Neither of these effects was observed; the patients demonstrated
N400 priming effects for both verb types, just as the controls
did.
The lack of a regular–irregular dissociation in the N400 data
of the patients, despite a dissociation in the response time data,
suggests that these two dependent measures do not reflect the
same cognitive processes. The absence of significant correlations
between the RT and N400 priming effect sizes also supports this
idea. The preservation of the regular-verb N400 priming effect
in the patients suggests that the lexical entries for the presenttense targets were successfully pre-activated by the past-tense
primes. Therefore, the absence of behavioral regular-verb priming
13
may not be attributable to a disruption of lexical access resulting from compromised morphological parsing mechanisms, as
argued by Marslen-Wilson and Tyler (2007), because this leads
to the prediction of reduced regular-verb N400 priming. Similarly, the absence of behavioral regular-verb priming may not be
attributable to changes in pre-lexical and lexical priming based
on ortho-phonological overlap, because this arguably also leads
to the prediction of reduced regular-verb N400 priming. Instead,
the absence of behavioral regular-verb priming may be due to
altered post-lexical processes following the N400. These postlexical processes might still be impacted by ortho-phonological
overlap between primes and targets, given the relationships in the
behavioral data between the priming of regular verbs and that of
the pseudopast and orthophono conditions.
The relative timing of the N400 and the behavioral response
can be accommodated in this hypothesis. For both controls
and patients, the differences between the N400 for primed and
unprimed items began to emerge as early as 350 ms following the
onset of the target words, which had a mean duration of 514 ms.
Thus, the N400 priming effect began before the target word had
been completely uttered. These effects resolved and returned to
the baseline provided by the unprimed ERPs by 850 ms (in the
case of the irregular verbs), if not sooner. In contrast, the grand
mean response times of 866 ms following word onset for the controls, and 1130 ms following word onset for the patients, suggest
that the behavioral response could reflect additional processing following the N400 and lexical access. A further suggestion from the
behavioral data that the lack of priming for regular verbs may have
arisen post-lexically is the significant relationship between participants’ mean response times and the size of the three priming
effects that were diminished in the patient group as a whole. The
longer the participants took, in general, to perform the lexical decision task, the smaller the priming effects were for regular verbs and
the two phonological control conditions (each p < 0.002). The size
of the irregular-verb priming effect was not correlated with mean
response time (p = 0.60).
The interpretation of the ERP data from the pseudopast and
orthophono conditions is less clear. In our study of young controls in this experimental design, we observed reduced N400 s
for prime-target pairs that shared word-onset ortho-phonological
overlap when both prime and target were presented in the auditory
modality (Justus et al., 2008), as in the current study. When presentation was cross-modal, with auditory primes and visual targets,
N400 effects were minimized and a post-lexical anterior negativity (PLAN), which affected the late positive component (LPC) of
the waveform, became more prominent (Justus et al., 2009). We
suggested that the PLAN might reflect a relatively late, strategic
process driven by word-onset ortho-phonological overlap between
primes and targets. If such post-lexical processes were disrupted
in the patients, as suggested above, this might lead to the prediction that that N400 effects would be preserved for the phonological
conditions, but PLAN modulations of the LPC would be altered.
We observed evidence for both N400 and PLAN effects in the current study, with differences between the controls and patients, and
between the pseudopast and orthophono conditions. Specifically,
during the N400 window, the controls demonstrated only a trend
for an effect, and only for the pseudopast condition (p = 0.07). In
contrast, the patients demonstrated significant effects for both the
pseudopast and orthophono conditions (each p < 0.01), indicating
that, like the controls of Justus et al. (2008), hearing a prime word
facilitated the lexical access of a target word with which it shared
word-onset ortho-phonological overlap. Why this effect would be
diminished in the age-matched controls, relative to the young controls of the earlier study, and yet be preserved in the patients is
unclear. Any effect of aging (see Federmeier & Kutas, 2005) would
be expected to apply to both groups of the current study.
14
During the LPC window, the controls demonstrated a marginal
reversal of the N400 effect over anterior electrodes (i.e., the PLAN),
but only for the pseudopast condition (p = 0.06). In contrast, the
patients demonstrated a continuation of the N400 effect, but only
for the orthophono condition (p = 0.04), while a weak trend for a
PLAN-direction reversal failed to approach significance. Given that
the N400 effects for primed items are more positive, and the PLAN
effects are more negative, the overlap between the two could cause
each one to cancel out the other. We speculate that this may be the
reason why the controls did not show significant N400 reductions
for the pseudopast and orthophono conditions, i.e., the already
evolving PLAN effects obscured them. In contrast, if PLAN effects
were diminished in the patients, this would make the remaining
N400 effects more prominent.
Despite the unresolved questions regarding the PLAN effects
on the late positive component, the ERP data clearly indicate preserved N400 priming for both regular and irregular verbs in both
LIFG patients and controls. This suggests that despite damage to
the LIFG, patients are able to pre-activate the lexical representations for regular present-tense forms (e.g., look) when they hear
the corresponding past-tense form (looked). This result is in direct
conflict with the notion of Marslen-Wilson and Tyler (2007) that,
when hearing a regular past-tense form, lexical access for the corresponding present tense is dependent upon an LIFG-dependent
mechanism to detect the regular inflectional affix and strip it from
the stem. The implications of our N400 results are either that affixstripping is not critically dependent on the LIFG, or, as argued by
single-system models, that affix-stripping is not necessary given
that all past-tense forms would be represented in the lexicon. The
result is also problematic for any single-system account that would
require the effects of ortho-phonological overlap in this experimental design to occur prior to lexical access.
5. Implications for theories of Broca’s area function
The discussion to this point has been framed in terms of two
competing views of Broca’s area function, as viewed from the perspective of the past-tense debate. Dual-system theories view the
function of this region as morphological. This view is aligned with
the notion that Broca’s area is needed for grammatical processing
more generally, including not only morphological but also syntactic
processing (Marslen-Wilson & Tyler, 1998; Pinker & Ullman, 2002;
Pinker, 1999). In contrast, single-system theories view the function of this region as phonological. This view is aligned with the
notion that Broca’s area is needed for speech processing, including not only speech production but also speech perception (Bird
et al., 2003; Joanisse & Seidenberg, 1999; McClelland & Patterson,
2002). Although our experiment was designed with the specific
predictions of single- and dual-system theories in mind, moving beyond the narrow focus of the past-tense debate may help
explain two unexpected findings from the present experiment:
(1) despite behavioral regular-verb priming being diminished in
the LIFG group, the corresponding N400 priming effects were
preserved, and (2) the expected relationships between the regularverb, pseudopast, and orthophono priming co-varied with overall
RT for the lexical decision task, such that patients who took longest
to respond tended to show reduced behavioral priming for these
three effects. Table 3 presents eleven hypotheses concerning the
function of Broca’s area – along with two hypotheses for the deficit
underlying non-fluent aphasia – and a prediction concerning how
disruption to each hypothesized function might impact the twoword past-tense priming design. The first two rows represent the
phonological and morphological views discussed thus far.
Although not typically cited by single-system authors, the view
that Broca’s aphasics have a generalized phonological deficit paral-
lels aspects of Blumstein and colleagues’ view that Broca’s aphasics
have a deficit in using phonological information to perform lexical
access. For instance, Broca’s aphasics showed a reduction in priming between semantically related words when the initial phoneme
of the prime word was phonetically altered (e.g., c*at-dog) or
replaced by another (e.g., gat-dog) (Aydelott Utman, Blumstein, &
Sullivan, 2001; Milberg, Blumstein, & Dworetzky, 1988; Misiurski,
Blumstein, Rissman, & Berman, 2005; for a review see Blumstein,
2007; also see Janse, 2006, 2008). Thus, regarding the present twoword priming design, this view might also predict a reduction
in word-onset ortho-phonological priming, which would account
for reductions in behavioral priming effects for the regular-verb,
pseudopast, and orthophono conditions, with relative preservation of irregular-verb priming. However, like Bird et al. (2003) and
Marslen-Wilson and Tyler (2007), this view would seem to attribute
phonological priming to pre-lexical events, and accordingly, would
also predict corresponding reductions in N400 priming, which was
instead preserved.
We next consider two proposed roles of Broca’s area in lexicalsemantic processing, reflecting neural events that occur during or
immediately following the N400. It has been proposed that the
LIFG, especially BA 44, is a candidate region for so-called mirror
neurons (Rizzolatti, Fogassi, & Gallese, 2001), though it has been
argued that evidence for this in humans is rather weak (Hickok,
2009). Nevertheless, this view might suggest that lexical activation
and associated lexical priming effects would be altered for verbs
that refer to imaginable actions, particularly actions performed by
the hand or arm. However, such items in our stimulus list were split
roughly equally between the regular (e.g., reached, placed, grabbed)
and irregular lists (e.g., dealt, threw, wrote), and would not seem to
account for a reduction in behavioral priming for the regular items
alone.
A strong case can be made for a role for Broca’s area in resolving
lexical-semantic ambiguity for word-forms that are associated with
more than one lexical entry (e.g., Rodd, Davis, & Johnsrude, 2005;
Davis et al., 2007; Zempleni, Renken, Hoeks, Hoogduin, & Stowe,
2007; Bedny, Hulbert, & Thompson-Schill, 2007). This might suggest that lexical activation and associated lexical priming effects for
ambiguous words would be altered. While we attempted to avoid
such items in designing our regular, pseudopast, and orthophono
conditions, a few items may be considered lexically ambiguous. In
contrast, many of the irregular past-tense items were unavoidably
lexically ambiguous (e.g., made, left, read). However, an interpretation based in lexical ambiguity would not seem to explain a
relative reduction in priming for the regular items, as observed in
the present study.
In addition to claims regarding morphological processing,
Broca’s area is frequently associated with syntactic processing
(see Sahin et al., 2006). Potential roles for the region in syntax include syntactic parsing (e.g., Friederici, 2002), syntactic
movement (Grodzinsky, 2000; Grodzinsky & Santi, 2008), or
the checking of thematic-role assignments (Caplan, Stanczak, &
Waters, 2008). Such syntactic roles, along with the literature
on semantic ambiguity, were integrated by Hagoort (2005) in a
memory-unification-control (MUC) framework. In this view, the
LIFG plays an essential role in unification, i.e., “the integration of
lexically retrieved information into a representation of multi-word
utterances” (2005, p. 416). While is it clear that syntactic roles, and
other views that entail the binding of words together in utterances,
need to be accounted for in any general theory of Broca’s area function, the predictions of these viewpoints for the two-word priming
design are unclear.
The final four rows of Table 3 represent proposed functions of
the LIFG that may potentially explain altered patterns of behavioral priming that are due to events following lexical access. Broca’s
area plays a role in verbal working memory, and has typically
15
Table 3
Theories of Broca’s area/LIFG function.
Role of Broca’s area/LIFG
Reference
Effect of disruption on past-tense priming
Phonological processing (perception and production)
Bird et al. (2003)
Morphological parsing
Marslen-Wilson and Tyler (2007)
Disrupted phonological representation, reduced
phonological priming
Disrupted segmentation of regular past-tense IRP into
stem and affix
Lexical activation, competition, and suppressiona
Action semantics
Blumstein (2007)
Janse (2006)
Rodd et al. (2005)
Bedny et al. (2007)
Rizzolatti et al. (2001)
Syntactic parsing
Syntactic movement
Thematic-role checking
Unification
Friederici (2002)
Grodzinsky (2000)
Caplan et al. (2008)
Hagoort (2005)
[No clear prediction for two-word priming]
Working memory, articulatory rehearsal
Paulesu et al. (1993)
Baldo and Dronkers (2006)
Hickok and Poeppel (2007)
Burton (2001, 2009)
Badre and Wagner (2007)
Gold et al. (2006)
Novick et al. (2005)
Thompson-Schill et al. (2005)
Disrupted covert articulation, reduced phonological
priming
Disrupted task strategies involving explicit sub-lexical
segmentation
Disrupted controlled components of priming and
lexical retrieval
Increased proactive interference between
phonologically similar primes and targets
a
Lexical-semantic ambiguity resolution
Speech production, explicit segmentation
Controlled retrieval
Representational conflict detection and resolution
a
Reduced lexical activation, reduced phonological
priming, increased interference
Altered priming for lexically ambiguous items
Altered priming for words describing imaginable
actions
Hypothesized deficit in nonfluent or Broca’s aphasia, which is not necessarily the result of damage to Broca’s area in the LIFG.
been associated with articulatory rehearsal within the Baddeley
(1986) phonological-loop framework (e.g., Baldo & Dronkers, 2006;
Paulesu, Frith, & Frackowiak, 1993). Some have argued that the role
of Broca’s area in syntactic processing may be best explained in
terms of verbal working memory (Kaan & Swaab, 2002; Rogalsky
& Hickok, in press). With regard to the current design, it is possible
that, while not required by the task instructions, healthy participants verbally rehearse the prime word in anticipation of the target.
Given the duration of the ISI between prime and target (on average
870 ms), this would likely amount to a single covert articulation of
the prime word. This could be the basis of at least part of the priming effects driven by phonological overlap that were observed in
the control group, but were absent in the LIFG patient group, due
to a diminished ability to articulate the prime word covertly during the interval between prime and target. In contrast, patterns of
activation within the mental lexicon would be unchanged, and give
rise to intact N400 priming effects in both groups.
A focus on covert articulation following lexical access is also
consistent with the speech comprehension model of Hickok and
Poeppel (2004, 2007). According to these authors, speech comprehension and lexical access in naturalistic situations depend upon a
ventral processing stream projecting through the temporal lobe.
Only speech perception tasks that have specific task demands,
such as the need for explicit sub-lexical segmentation (see Burton,
2001, 2009) will recruit a dorsal stream that projects to inferior
frontal regions, and which is also used during speech production.
From the perspective of the Hickok–Poeppel model, neither the
dual-system notion that the LIFG is necessary for parsing heard
inflectional morphemes (Marslen-Wilson & Tyler, 2007), nor the
single-system notion that the LIFG is necessary for phonological
aspects of spoken-language processing (Bird et al., 2003) would
be predicted. Instead, this view might suggest that any effects
of LIFG damage on performance in an auditory lexical priming
task would be due to task demands and explicit strategies. For
instance, healthy participants may become consciously aware of
the morpho-phonological relationships between primes and targets, and develop a strategy to anticipate likely targets by explicitly
segmenting the final phonemes of the prime words (see McQueen
& Sereno, 2005), recruiting the articulatory mechanisms of the LIFG
(see Burton, 2001, 2009). Such a strategy would contribute to prim-
ing for the regular, pseudopast, and orthophono conditions, and
to a lesser extent, the weak-irregular condition. Given their problems with covert articulation, people with LIFG damage would not
benefit from such segmentation strategies, which would impact
response times for these conditions. Again, automatic lexical priming within the temporal lobe would be intact, accounting for the
preserved N400 priming effects.
The final two theories listed in Table 3 also may be able
to account for changes in behavioral priming related to overall
RT and accompanied by preserved N400 priming. These theories
argue for a more domain-general role in cognitive control for
the LIFG, which in this literature is often referred to as the ventrolateral prefrontal cortex (VLPFC). One such view is that the
LIFG, particularly the pars orbitalis (BA 47), supports controlled
retrieval of goal-relevant representations (Badre & Wagner, 2007).
For instance, the LIFG has been implicated in two-word semantic
priming designs only when the interval between prime and target is relatively long (e.g., 1000 ms), arguably permitting strategic
retrieval in addition to automatic lexical activation (Gold et al.,
2006). This finding echoes an earlier debate in the patient literature concerning whether the performance of Broca’s aphasics in
these tasks is better explained in terms of altered automatic or
controlled processes (see Hagoort, 1997; Milberg, Blumstein, Katz,
Gershberg, & Brown, 1995; Tyler, Ostrin, Cooke, & Moss, 1995). The
notion of controlled retrieval is also particularly relevant for the
role of Broca’s area in the past-tense production task described
in the Introduction; Broca’s area is recruited for the production
of both regular and irregular past tenses, and especially so for
irregulars. From a single-system perspective, the less consistent
phonological mapping between irregular present- and past-tense
forms might be expected to lead to a greater need for controlled
retrieval (Seidenberg & Arnoldussen, 2003). From a dual-system
perspective, greater cognitive control might be expected for the
production of irregular past tenses, because the speaker may be
inhibiting an over-regularization (Beretta, Campbell, et al., 2003).
Another hypothesis regarding the LIFG/VLPFC in cognitive control is that this region detects and resolves conflict between
competing representations (Novick, Trueswell, & Thompson-Schill,
2005; Thompson-Schill, Bedny, & Goldberg, 2005). In the twoprocess view of Badre and Wagner (2007), this role is ascribed to
16
the pars triangularis (BA 45). The conflict-resolution view is notable
in that it integrates a variety of seemingly disparate results into
a unified view of LIFG function; findings in phonological ambiguity, semantic ambiguity, and syntactic ambiguity are grounded
in a larger literature implicating the LIFG in representationalinterference designs such as the Stroop (January, Trueswell, &
Thompson-Schill, 2009). For instance, in the Sternberg itemrecognition task, participants view a group of letters and then must
respond to a probe item to indicate whether it was part of the
immediately preceding set. In a modification designed to create
proactive interference, representational and response conflict can
be manipulated depending on whether the target item was also a
part of the memory set of the preceding trial, leading to increased
recruitment of the pars triangularis (Jonides & Nee, 2006). Some
patients with focal lesion to the LIFG have difficulty specifically with
these trials (Thompson-Schill et al., 2002; Novick, Kan, Trueswell,
& Thompson-Schill, 2009). In the current study, we speculate that
representational conflict may occur when a prime-target pair that
differs by a single phoneme is presented. Healthy controls may have
no difficulty inhibiting or directing attention away from the representation of the prime in order to respond to the target, regardless
of the phonological similarity of the two items. In contrast, patients
with impaired cognitive control may have difficulty with phonologically similar primes and targets due to a representational conflict
between the two.
Two results of the current study, while being the most unexpected with regard to the predictions of dual- and single-system
models, are intriguing in light of these last four views of LIFG function. The first is the observation that the absence of regular-verb
priming in the patients’ behavioral data was not accompanied by
any reduction in N400 priming, a result which might have been
expected from both dual- and single-system perspectives. The second observation is the significant relationship between the three
priming effects that were abolished in the LIFG patient group (i.e.,
priming for regular verbs, pseudopast, and orthophono) and the
amount of time in general a participant required to perform the
lexical decision task, which is not easily accounted for by either
dual- or single-system views. Both of these results suggest that the
“deficit” that the LIFG patients have with the regular past tense
and the phonological control conditions is not intrinsic to the perception of these word forms, but rather is an effect that emerges
following lexical access in this particular experimental design. As
our brief review indicates, these effects instead could be based
in previously well documented functions of the LIFG in speech
production (covert articulation, segmentation strategies) and/or
cognitive control (controlled retrieval, representational conflict).
6. Conclusion
Here we replicated a previously reported dissociation between
disrupted regular-verb priming and preserved irregular-verb priming following brain damage. That our eleven patients were chosen
based on documented damage to Broca’s area strengthens the argument that this dissociation is a consequence of damage to the LIFG.
Examination of the behavioral data revealed that absent regularverb priming covaried with absent ortho-phonological priming,
regardless of whether phonological cues to morphological structure were present. Further, a division of stimuli into regular, weak
irregular, and strong verbs suggested that the regular–irregular
dissociation was continuous, rather than categorical. Both of
these results are predicted by the single-system, connectionist
approach to inflectional morphology and are not predicted by
current dual-system, rule-based models. Event-related potential
data demonstrated that N400 priming effects were intact, suggesting that the absence of significant regular-verb priming in the
behavioral data of these patients was not the result of impairments in pre-lexical or lexical priming, and was instead the result
of post-lexical events related to covert articulation, segmentation
strategies, and/or cognitive control.
Acknowledgements
For their assistance with various stages of this project, we thank
Doug Mobley, Randi Johnsen, Adrian Willoughby, And Turken,
David Wilkins, and Alex Hurst. This research was funded by NIH
grant P01NS40813 and was supported by the resources and facilities of the Dept. of Veterans Affairs Northern California Health Care
System in Martinez, CA. Some of these data were previously presented at the 2007 meeting of the Cognitive Neuroscience Society
in New York.
References
Aydelott Utman, J., Blumstein, S. E., & Sullivan, K. (2001). Mapping from sound to
meaning: Reduced lexical activation in Broca’s aphasics. Brain and Language, 79,
444–472.
Baayen, R. H., Piepenbrock, R., & Gulikers, L. (1995). The CELEX lexical database (CDROM). Philadelphia, PA: Linguistic Data Consortium, University of Pennsylvania.
Baddeley, A. (1986). Working memory. Oxford: Oxford University Press.
Badre, D., & Wagner, A. D. (2007). Left ventrolateral prefrontal cortex and the cognitive control of memory. Neuropsychologia, 45, 2883–2901.
Baldo, J. V., & Dronkers, N. F. (2006). The role of inferior parietal and inferior frontal
cortex in working memory. Neuropsychology, 20, 529–538.
Bates, E., Wilson, S. M., Saygin, A. P., Dick, F., Sereno, M. I., Knight, R. T., et al. (2003).
Voxel-based lesion-symptom mapping. Nature Neuroscience, 6, 448–450.
Bedny, M., Hulbert, J. C., & Thompson-Schill, S. L. (2007). Understanding words in
context: The role of Broca’s area in word comprehension. Brain Research, 1146,
101–114.
Bentin, S., & Peled, B.-S. (1990). The contribution of task-related factors to ERP repetition effects at short and long lags. Memory & Cognition, 18, 359–366.
Beretta, A., Campbell, C., Carr, T. H., Huang, J., Schmitt, L. M., Christianson, K., et al.
(2003). An ER-fMRI investigation of morphological inflection in German reveals
that the brain makes a distinction between regular and irregular forms. Brain
and Language, 85, 67–92.
Beretta, A., Carr, T. H., Huang, J., & Cao, Y. (2003). The brain is not single-minded about
inflectional morphology: A response to the commentaries. Brain and Language,
85, 531–534.
Bird, H., Lambon Ralph, M. A., Seidenberg, M. S., McClelland, J. L., & Patterson, K.
(2003). Deficits in phonology and past-tense morphology: What’s the connection? Journal of Memory and Language, 48, 502–526.
Blumstein, S. (2007). Word recognition in aphasia. In M. G. Gaskell, & G. Altmann
(Eds.), The Oxford handbook of psycholinguistics (pp. 141–155). Oxford: Oxford
University Press.
Braber, N., Patterson, K., Ellis, K., & Lambon Ralph, M. A. (2005). The relationship
between phonological and morphological deficits in Broca’s aphasia: Further
evidence from errors in verb inflection. Brain and Language, 92, 278–287.
Burton, M. W. (2001). The role of inferior frontal cortex in phonological processing.
Cognitive Science, 25, 695–709.
Burton, M. W. (2009). Understanding the role of the prefrontal cortex in phonological
processing. Clinical Linguistics and Phonetics, 23, 180–195.
Caplan, D., Stanczak, L., & Waters, G. (2008). Syntactic and thematic constraint effects
on blood oxygenation level dependent signal correlates of comprehension of
relative clauses. Journal of Cognitive Neuroscience, 20, 643–656.
Davis, M. H., Coleman, M. R., Absalom, A. R., Rodd, J. M., Johnsrude, I. S., Matta,
B. F., et al. (2007). Dissociating speech perception and comprehension at
reduced levels of awareness. Proceedings of the National Academy of Sciences,
41, 16032–16037.
de Diego Balaguer, R., Rodríguez-Fornells, A., Rotte, M., Bahlmann, J., Heinze, H.J., & Münte, T. F. (2006). Neural circuits subserving the retrieval of stems and
grammatical features in regular and irregular verbs. Human Brain Mapping, 27,
874–888.
Desai, R., Conant, L. L., Waldron, E., & Binder, J. R. (2006). fMRI of past tense processing: The effects of phonological complexity and task difficulty. Journal of
Cognitive Neuroscience, 18, 278–297.
Dhond, R. P., Marinkovic, K., Dale, A. M., Witzel, T., & Halgren, E. (2003). Spatiotemporal maps of past-tense verb inflection. Neuroimage, 19, 91–100.
Dronkers, N. F., & Baldo, J. V. (2009). Language: Aphasia. In L. R. Squire (Ed.), Encyclopedia of neuroscience (pp. 343–348). Oxford: Academic Press.
Dronkers, N. F., Plaisant, O., Iba-Zizen, M. T., & Cabanis, E. A. (2007). Paul Broca’s
historic cases: High resolution MR imaging of the brains of Leborgne and Lelong.
Brain, 130, 1432–1441.
Faroqi-Shah, Y. (2007). Are regular and irregular verbs dissociated in non-fluent
aphasia? A meta-analysis. Brain Research Bulletin, 74, 1–13.
Federmeier, K. D. (2007). Thinking ahead: The role and roots of prediction in language comprehension. Psychophysiology, 44, 491–505.
Federmeier, K. D., & Kutas, M. (2005). Aging in context: Age-related changes in
context use during language comprehension. Psychophysiology, 42, 133–141.
Fowler, C. A., Napps, S. E., & Feldman, L. (1985). Relations among regular and irregular
morphologically related words in the lexicon as revealed by repetition priming.
Memory & Cognition, 13, 241–255.
Friederici, A. D. (2002). Towards a neural basis of auditory sentence processing.
Trends in Cognitive Sciences, 6, 78–84.
Friederici, A. D., Hahne, A., & von Cramon, D. Y. (1998). First-pass versus second-pass
parsing processes in a Wernicke’s and a Broca’s aphasic: Electrophysiological
evidence for a double dissociation. Brain and Language, 62, 311–341.
Friederici, A. D., von Cramon, D. Y., & Kotz, S. A. (1999). Language related brain potentials in patients with cortical and subcortical left hemisphere lesions. Brain, 122,
1033–1047.
Gold, B. T., Balota, D. A., Jones, S. J., Powell, D. K., Smith, C. D., & Andersen, A. H.
(2006). Dissociation of automatic and strategic lexical-semantics: Functional
magnetic resonance imaging evidence for differing roles of multiple frontotemporal regions. Journal of Neuroscience, 26, 6523–6532.
Gonnerman, L. M., Seidenberg, M. S., & Andersen, E. S. (2007). Graded semantic and
phonological similarity effects in priming: Evidence for a distributed connectionist approach to morphology. Journal of Experimental Psychology: General, 136,
323–345.
Grodzinsky, Y. (2000). The neurology of syntax: Language use without Broca’s area.
Behavioral and Brain Sciences, 23, 1–71.
Grodzinsky, Y., & Santi, A. (2008). The battle for Broca’s region. Trends in Cognitive
Sciences, 12, 474–480.
Gross, M., Say, T., Kleingers, M., Clahsen, H., & Münte, T. F. (1998). Human brain
potentials to violations in morphologically complex Italian words. Neuroscience
Letters, 241, 83–86.
Hagoort, P. (1997). Semantic priming in Broca’s aphasics at a short SOA: No support
for an automatic access deficit. Brain and Language, 56, 287–300.
Hagoort, P. (2005). On Broca, brain, and binding: A new framework. Trends in Cognitive Sciences, 9, 416–423.
Hagoort, P. (2008). The fractionation of spoken language understanding by measuring electrical and magnetic brain signals. Philosophical Transactions of the Royal
Society B, 363, 1055–1069.
Hagoort, P., Brown, C., & Swaab, T. Y. (1996). Lexical-semantic event-related potential effects in patients with left hemisphere lesions and aphasia, and patients
with right hemisphere lesions without aphasia. Brain, 119, 627–649.
Hanson, V. L., & Wilkenfeld, D. (1985). Morphophonology and lexical organization
in deaf readers. Language and Speech, 28, 269–279.
Hay, J. B., & Baayen, R. H. (2005). Shifting paradigms: Gradient structure in morphology. Trends in Cognitive Sciences, 9, 342–348.
Hickok, G. (2009). Eight problems for the mirror neuron theory of action understanding in monkeys and humans. Journal of Cognitive Neuroscience, 21, 1229–1243.
Hickok, G., & Poeppel, D. (2004). Dorsal and ventral streams: A framework for understanding aspects of the functional anatomy of language. Cognition, 92, 67–99.
Hickok, G., & Poeppel, D. (2007). The cortical organization of speech processing.
Nature Reviews Neuroscience, 8, 393–402.
Holcomb, P. J., Anderson, J., & Grainger, J. (2005). An electrophysiological study of
cross-modal repetition priming. Psychophysiology, 42, 493–507.
Holcomb, P. J., Grainger, J., & O’Rourke, T. (2002). An electrophysiological study of the
effects of orthographic neighborhood size on printed word perception. Journal
of Cognitive Neuroscience, 14, 938–950.
Jaeger, J. J. (2003). Commentary on A Beretta et al. ‘An ER-fMRI investigation of
morphological inflection in German reveals that the brain makes a distinction
between regular and irregular forms’. Brain and Language, 85, 524–526.
Jaeger, J. J., Lockwood, A. H., Kemmerer, D. L., Van Valin, R. D., Jr., Murphy, B. W.,
& Khalak, H. G. (1996). A positron emission tomographic study of regular and
irregular verb morphology in English. Language, 72, 451–497.
Jaeger, J. J., Van Valin, R. D., Jr., & Lockwood, A. H. (1998). Reply to Seidenberg and
Hoeffner. Language, 74, 123–128.
Janse, E. (2006). Lexical competition effects in aphasia: Deactivation of lexical candidates in spoken word processing. Brain and Language, 97, 1–11.
Janse, E. (2008). Spoken-word processing in aphasia: Effects of item overlap and
item repetition. Brain and Language, 105, 185–198.
January, D., Trueswell, J. C., & Thompson-Schill, S. L. (2009). Co-localization of Stroop
and syntactic ambiguity resolution in Broca’s area: Implications for the neural basis of sentence processing. Journal of Cognitive Neuroscience, 21, 2434–
2444.
Jasper, H. H. (1958). The ten-twenty electrode system of the International Federation.
Electroencephalography and Clinical Neurophysiology, 10, 371–375.
Joanisse, M. F., & Seidenberg, M. S. (1999). Impairments in verb morphology after
brain injury: A connectionist model. Proceedings of the National Academy of Sciences, 96, 7592–7597.
Joanisse, M. F., & Seidenberg, M. S. (2005). Imaging the past: Neural activation in
frontal and temporal regions during regular and irregular past-tense processing.
Cognitive, Affective, & Behavioral Neuroscience, 5, 282–296.
Jonides, J., & Nee, D. E. (2006). Brain mechanisms of proactive interference in working
memory. Neuroscience, 139, 181–193.
Justus, T., Larsen, J., de Mornay Davies, P., & Swick, D. (2008). Interpreting dissociations between regular and irregular past-tense morphology: Evidence
from event-related potentials. Cognitive, Affective, & Behavioral Neuroscience, 8,
178–194.
Justus, T., Yang, J., Larsen, J., de Mornay Davies, P., & Swick, D. (2009). An event-related
potential study of cross-modal morphological and phonological priming. Journal
of Neurolinguistics, 22, 584–604.
17
Kaan, E., & Swaab, T. Y. (2002). The brain circuitry of syntactic comprehension. Trends
in Cognitive Sciences, 6, 350–356.
Keller, S. S., Crow, T., Foundas, A., Amunts, K., & Roberts, N. (2009). Broca’s area:
Nomenclature, anatomy, typology and asymmetry. Brain and Language, 109,
29–48.
Kielar, A., & Joanisse, M. F. (2010). Graded effects of regularity in language revealed
by N400 indices of morphological priming. Journal of Cognitive Neuroscience, 22,
1373–1398.
Kielar, A., Joanisse, M. F., & Hare, M. L. (2008). Priming English past tense verbs: Rules
or statistics? Journal of Memory and Language, 58, 327–346.
Kutas, M., & Federmeier, K. D. (2000). Electrophysiology reveals semantic memory
use in language comprehension. Trends in Cognitive Sciences, 4, 463–470.
Kutas, M., & Hillyard, S. A. (1980). Reading senseless sentences: Brain potentials
reflect semantic incongruity. Science, 207, 203–205.
Lambon Ralph, M. A., Braber, N., McClelland, J. L., & Patterson, K. (2005). What
underlies the neuropsychological pattern of irregular > regular past-tense verb
production? Brain and Language, 93, 106–119.
Larsen, J., Justus, T., de Mornay Davies, P., & Swick, D. (2008). No dissociation in
the production of regular and irregular past-tense verbs in left frontal patients.
Journal of Cognitive Neuroscience Supplement, 106.
Lau, E. F., Phillips, C., & Poeppel, D. (2008). A cortical network for semantics:
(De)constructing the N400. Nature Reviews Neuroscience, 9, 920–933.
Lau, E., Almeida, D., Hines, P. C., & Poeppel, D. (2009). A lexical basis for N400 context
effects: Evidence from MEG. Brain and Language, 111, 161–172.
Longworth, C. E., Keenan, S. E., Barker, R. A., Marslen-Wilson, W. D., & Tyler, L.
K. (2005). The basal ganglia and rule-governed language use: Evidence from
vascular and degenerative conditions. Brain, 128, 584–596.
Longworth, C. E., Marslen-Wilson, W. D., Randall, B., & Tyler, L. K. (2005). Getting to
the meaning of the regular past tense: Evidence from neuropsychology. Journal
of Cognitive Neuroscience, 17, 1087–1097.
Marslen-Wilson, W. D. (2007). Morphological processes in language comprehension.
In M. G. Gaskell, & G. Altmann (Eds.), The Oxford handbook of psycholinguistics
(pp. 175–193). Oxford: Oxford University Press.
Marslen-Wilson, W. D., & Tyler, L. K. (1997). Dissociating types of mental computation. Nature, 387, 592–594.
Marslen-Wilson, W. D., & Tyler, L. K. (2007). Morphology, language and the brain: The
decompositional substrate for language comprehension. Philosophical Transactions of the Royal Society B, 362, 823–836.
Marslen-Wilson, W., & Tyler, L. K. (1998). Rules, representations, and the English
past tense. Trends in Cognitive Sciences, 2, 428–435.
McClelland, J. L., & Patterson, K. (2002). Rules or connections in past-tense inflections: What does the evidence rule out? Trends in Cognitive Sciences, 6, 465–472.
McQueen, J. M., & Sereno, J. (2005). Cleaving automatic processes from strategic
biases in phonological priming. Memory & Cognition, 33, 1185–1209.
Milberg, W., Blumstein, S. E., Katz, D., Gershberg, F., & Brown, T. (1995). Semantic facilitation in aphasia: Effects of time and expectancy. Journal of Cognitive
Neuroscience, 7, 33–50.
Milberg, W., Blumstein, S., & Dworetzky, B. (1988). Phonological processing and
lexical access in aphasia. Brain and Language, 34, 279–293.
Miozzo, M. (2003). On the processing of regular and irregular forms of verbs and
nouns: Evidence from neuropsychology. Cognition, 87, 101–127.
Miozzo, M., & Gordon, P. (2005). Facts, events, and inflection: When language and
memory dissociate. Journal of Cognitive Neuroscience, 17, 1074–1086.
Misiurski, C., Blumstein, S. E., Rissman, J., & Berman, D. (2005). The role of lexical competition and acoustic-phonetic structure in lexical processing: Evidence
from normal subjects and aphasic patients. Brain and Language, 93, 64–78.
Morris, J., & Holcomb, P. J. (2005). Event-related potentials to violations of inflectional verb morphology in English. Cognitive Brain Research, 25, 963–981.
Münte, T. F., Say, T., Clahsen, H., Schiltz, K., & Kutas, M. (1999). Decomposition of
morphologically complex words in English: Evidence from event-related brain
potentials. Cognitive Brain Research, 7, 241–253.
Nagy, M. E., & Rugg, M. D. (1989). Modulation of event-related potentials by word
repetition: The effects of inter-item lag. Psychophysiology, 26, 431–436.
Napps, S. E. (1989). Morphemic relationships in the lexicon: Are they distinct from
semantic and formal relationships? Memory & Cognition, 17, 729–739.
Newman, A. J., Ullman, M. T., Pancheva, R., Waligura, D. L., & Neville, H. (2007).
An ERP study of regular and irregular English inflection. Neuroimage, 34, 435–
445.
Novick, J. M., Kan, I. P., Trueswell, J. C., & Thompson-Schill, S. L. (2009). A case for
conflict across multiple domains: Memory and language impairments following damage to ventrolateral prefrontal cortex. Cognitive Neuropsychology, 26,
527–567.
Novick, J. M., Trueswell, J. C., & Thompson-Schill, S. L. (2005). Cognitive control
and parsing: Reexamining the role of Broca’s area in sentence comprehension.
Cognitive, Affective, & Behavioral Neuroscience, 5, 263–281.
Olichney, J. M., Van Petten, C., Paller, K. A., Salmon, D. P., Iragui, V. J., & Kutas, M.
(2000). Word repetition in amnesia: Electrophysiological measures of impaired
and spared memory. Brain, 123, 1948–1963.
Patterson, K., Lambon Ralph, M. A., Hodges, J. R., & McClelland, J. L. (2001). Deficits
in irregular past-tense verb morphology associated with degraded semantic
knowledge. Neuropsychologia, 39, 709–724.
Paulesu, E., Frith, C. D., & Frackowiak, R. S. J. (1993). The neural correlates of the
verbal component of working memory. Nature, 362, 342–345.
Penke, M., Weyerts, H., Gross, M., Zander, E., Münte, T. F., & Clahsen, H. (1997). How
the brain processes complex words: An event-related potential study of German
verb inflections. Cognitive Brain Research, 6, 37–52.
18
Perrin, F., Pernier, J., Bertrand, O., & Echallier, J. F. (1989). Spherical splines for scalp
potential and current density mapping. Electroencepalography and Clinical Neurophysiology, 72, 184–187.
Pinker, S. (1999). Words and rules: The ingredients of language. New York: Basic.
Pinker, S., & Ullman, M. T. (2002). The past and future of the past tense. Trends in
Cognitive Sciences, 6, 456–463.
Post, B., Marslen-Wilson, W. D., Randall, B., & Tyler, L. K. (2008). The processing
of English regular inflections: Phonological cues to morphological structure.
Cognition, 109, 1–17.
Rastle, K., Harrington, J., & Coltheart, M. (2002). 358,534 nonwords: The ARC nonword database. Quarterly Journal of Experimental Psychology, 55A, 1339–1362.
Rizzolatti, G., Fogassi, L., & Gallese, V. (2001). Neurophysiological mechanisms
underlying the understanding and imitation of action. Nature Reviews Neuroscience, 2, 661–670.
Rodd, J. M., Davis, M. H., & Johnsrude, I. S. (2005). The neural mechanisms of
speech comprehension: fMRI studies of semantic ambiguity. Cerebral Cortex, 15,
1261–1269.
Rodríguez-Fornells, A., Clahsen, H., Lleó, C., Zaake, W., & Münte, T. F. (2001). Eventrelated brain responses to morphological violations in Catalan. Cognitive Brain
Research, 11, 47–58.
Rodríguez-Fornells, A., Münte, T. F., & Clahsen, H. (2002). Morphological priming
in Spanish verb forms: An ERP repetition priming study. Journal of Cognitive
Neuroscience, 14, 443–454.
Rogalsky, C., & Hickok, G. (in press). The role of Broca’s area in sentence comprehension. Journal of Cognitive Neuroscience.
Rorden, C., & Brett, M. (2000). Stereotaxic display of brain lesions. Behavioural Neurology, 12, 191–200.
Rugg, M. D. (1985). The effects of semantic priming and word repetition on eventrelated potentials. Psychophysiology, 22, 642–647.
Sach, M., Seitz, R. J., & Indefrey, P. (2004). Unified inflectional processing of regular
and irregular verbs: A PET study. Neuroreport, 15, 533–537.
Sahin, N. T., Pinker, S., & Halgren, E. (2006). Abstract grammatical processing of nouns
and verbs in Broca’s area: Evidence from fMRI. Cortex, 42, 540–562.
Seidenberg, M. S., & Arnoldussen, A. (2003). The brain makes a distinction between
hard and easy stimuli: Comments on Beretta et al. Brain and Language, 85,
527–530.
Seidenberg, M. S., & Hoeffner, J. H. (1998). Evaluating behavioral and neuroimaging
data on past tense processing. Language, 74, 104–122.
Sonnenstuhl, I., Eisenbeiss, S., & Clahsen, H. (1999). Morphological priming in the
German mental lexicon. Cognition, 72, 203–236.
Stanners, R. F., Neiser, J. J., Hernon, W. P., & Hall, R. (1979). Memory representation
for morphologically related words. Journal of Verbal Learning and Verbal Behavior,
18, 399–412.
Swaab, T., Brown, C., & Hagoort, P. (1997). Spoken sentence comprehension in aphasia: Event-related potential evidence for a lexical integration deficit. Journal of
Swaab, T., Brown, C., & Hagoort, P. (1998). Understanding ambiguous words in sentence contexts: Electrophysiological evidence for delayed contextual selection
in Broca’s aphasia. Neuropsychologia, 36, 737–761.
Swick, D., Kutas, M., & Knight, R. T. (1998). Prefrontal lesions eliminate the LPC but
do not affect the N400 during sentence reading. Journal of Cognitive Neuroscience
Supplement, 29.
Thompson-Schill, S. L., Bedny, M., & Goldberg, R. F. (2005). The frontal lobes and
the regulation of mental activity. Current Opinion in Neurobiology, 15, 219–
224.
Thompson-Schill, S. L., Jonides, J., Marshuetz, C., Smith, E. E., D’Esposito, M., Kan, I.
P., et al. (2002). Effects of frontal lobe damage on interference effects in working
memory. Cognitive, Affective, & Behavioral Neuroscience, 2, 109–120.
Tyler, L. K., de Mornay Davies, P., Anokhina, R., Longworth, C., Randall, B., &
Marslen-Wilson, W. D. (2002). Dissociations in processing past tense morphology: Neuropathology and behavioral studies. Journal of Cognitive Neuroscience,
14, 79–94.
Tyler, L. K., Randall, B., & Marslen-Wilson, W. D. (2002). Phonology and neuropsychology of the English past tense. Neuropsychologia, 40, 1154–1166.
Tyler, L. K., Marslen-Wilson, W. D., & Stamatakis, E. A. (2005). Differentiating lexical
form, meaning, and structure in the neural language system. Proceedings of the
National Academy of Sciences, 102, 8375–8380.
Tyler, L. K., Stamatakis, E. A., Post, B., Randall, B., & Marslen-Wilson, W. (2005). Temporal and frontal systems in speech comprehension: An fMRI study of past tense
processing. Neuropsychologia, 43, 1963–1974.
Tyler, L. K., Ostrin, R. K., Cooke, M., & Moss, H. E. (1995). Automatic access of lexical
information in Broca’s aphasics: Against the automaticity hypothesis. Brain and
Language, 48, 131–162.
Tyler, L. K., Stamatakis, E. A., Jones, R. W., Bright, P., Acres, K., & Marslen-Wilson, W. D.
(2004). Deficits for semantics and the irregular past tense: A causal relationship?
Journal of Cognitive Neuroscience, 16, 1159–1172.
Ullman, M. T., Corkin, S., Coppola, M., Hickok, G., Growdon, J. H., Koroshetz, W. J., et al.
(1997). A neural dissociation within language: Evidence that the mental dictionary is part of declarative memory, and that grammatical rules are processed by
the procedural system. Journal of Cognitive Neuroscience, 9, 266–276.
Ullman, M. T, Pancheva, R., Love, T., Yee, E., Swinney, D., & Hickok, G. (2005). Neural
correlates of lexicon and grammar: Evidence from the production, reading, and
judgment of inflection in aphasia. Brain and Language, 93, 185–238.
Van Petten, C., & Kutas, M. (1990). Interactions between sentence context and
word frequency in event-related brain potentials. Memory & Cognition, 18, 380–
393.
Van Petten, C., & Luka, B. J. (2006). Neural localization of semantic context effects
in electromagnetic and hemodynamic studies. Brain and Language, 97, 279–
293.
Van Petten, C., Macizo, P., & O’Rourke, P. (2007). Access to embedded words in speech
perception. Psychophysiology, 44, S5.
West, W. C., & Holcomb, P. J. (2000). Imaginal, semantic, and surface-level processing
of concrete and abstract words: An electrophysiological investigation. Journal of
Weyerts, H., Münte, T. F., Smid, H. G. O. M., & Heinze, H.-J. (1996). Mental representations of morphologically complex words: An event-related potential study with
adult humans. Neuroscience Letters, 206, 125–128.
Weyerts, H., Penke, M., Dohrn, U., Clahsen, H., & Münte, T. F. (1997). Brain potentials
indicate differences between regular and irregular German plurals. Neuroreport,
8, 957–962.
Zempleni, M.-Z., Renken, R., Hoeks, J. C. J., Hoogduin, J. M., & Stowe, L. A. (2007).
Semantic ambiguity processing in sentence context: Evidence from eventrelated fMRI. Neuroimage, 34, 1270–1279.

The role of Broca`s area in regular past-tense morphology

Transcription

Similar documents

Olivia Fiske, Alix Alto, Olivia Dooley, Dana Walker, Nicole Betz, Lisa

english - Ganatleba

Technical Data Sheet

kilz® original interior oil-based primer

KILZ COMPLETE® INTERIOR / EXTERIOR OIL